Center for Effective Organizations [PDF]

Center for Effective Organizations

STRATEGIC REWARD SYSTEMS CEO PUBLICATION G 92-2 (205) EDWARD E. L AWLER III University of Southern California G. DOUGLAS JENKINS Arizona State University

January, 1992

In M.D. Dunnette & L.M. Hough (Eds.), Handbook of Industrial & Organizational Psychology, 2 nd Edition, Palo Alto, CA: Consulting Psychologist Press Inc. (forthcoming)

C e n t e r for Ef f e c t i v e O r g a n i z a t i o n s - M a r s h a l l S c h o o l o f B u s i n e ss U n i v e r s i t y o f S o u t h e r n C a l i f o r n i a - L o s A n g e l e s, C A 9 0 0 8 9 – 0 8 0 6 TEL (2 1 3) 7 4 0 - 9 8 1 4 FAX (2 1 3) 7 4 0 - 4 3 5 4

http://www.marshall.usc.edu/ceo

STRATEGIC REWARD SYSTEMS

CEO Publication T 92-2 (205) Edward E. Lawler III University of Southern California G. Douglas Jenkins University of Arkansas Management Department Building BADM 402 Fayetteville, Arkansas 72701 (501) 575-6227

In M.D. Dunnette & L.M. Hough (Eds.), Handbook of Industrial & Organizational Psychology, 2nd Edition, Palo Alto, CA: Consulting Psychologist Press Inc. (forthcoming)

January, 1992 Center for Effective Organizations School of Business Administration University of Southern California Los Angeles, CA 90089-1421 (213) 743-8765

Abstract This chapter presents theory and research on the role of pay systems in complex organizations. It first looks at the relationship between pay systems and business strategy and organization structure. It then focuses on the impact of pay systems on six important determinants of organizational effectiveness: performance motivation, skill development motivation, attraction and retention, organization structure, culture, and costs. Different approaches to pay are reviewed with a focus on their impact on these six determinants of organizational effectiveness. The specific pay system practices reviewed include job-based pay, personbased pay, individual pay for performance systems, organizational pay for performance systems, and executive compensation pay practices. Overall the evidence indicates that pay systems have a strong impact on organizational effectiveness and there is a link between strategy, structure, and pay systems.

Rewards systems and their role in organizations have been studied from many perspectives and by multiple disciplines. Psychology, sociology, anthropology, and economics all have contributed to the literature. This chapter will focus on the design features of the major organizational reward system, the pay system, and its relationship to organizational behavior. Since this topic is not the province of a single discipline, the literature will be drawn from a variety of sources. The underlying assumption is that reward systems have wide ranging impacts on organizations and that their impact is greatly affected by their design and by the organizational context in which they operate. Thus to understand pay systems in organizations it is necessary to focus on the characteristics of both the organization and the pay system. Pay systems are one of the most prominent and important features of any organization. There is an extensive literature in psychology on the effects of rewards on human behavior that can help the understanding of the impact of reward systems. There is also considerable research on the impact of pay systems on individual behavior in organizations and on the behavior of organizations. Some areas, such as the design of job evaluation systems and the impact of merit pay plans, have been extensively researched, but many others have received little attention. The discussion in this chapter is intended to accomplish two purposes, first to provide an overview of the research which has been done and second to identify those areas which are important but which have been neglected by researchers.

Reward Systems and Strategy There is a growing body of literature on the relationship between organization strategy and human resource management systems (see e.g., Lawler, 1986a; Ulrich & Lake, 1990). It typically suggests that once its strategy is developed an organization needs to focus on the kind of human resources, culture, and behavior that is needed in order to make it effective. The next step is to design reward and other human

3

resource management systems that will motivate the right kind of performance, attract the right kinds of people, and create a supportive climate and structure (see e.g., Balkin & Gomez-Mejia, 1987b; Hambrick & Snow, 1989; Lawler, 1986b). Much of the writing on strategy is conceptual and descriptive in nature. It starts from a competitive advantage framework and focuses on how human resource management practices can enhance an organization's performance by supporting particular kinds of behavior (see e.g., Porter, 1985, 1990; Ulrich & Lake, 1990). Depending on the actual practices that are recommended there may or may not be evidence that the practices are used and can produce the desired results. For example, if an organization is trying to become a low cost producer it is often argued that incentive plans and other pay for performance plans are called for. Missing is evidence that companies in fact are using plans with particular strategic agendas in mind. Only in the area of management compensation are there studies that attempt to link strategic agendas and compensation practices (see e.g., Hambrick & Snow, 1989; Kerr, 1985). There is another way in which the reward system may need to be taken into consideration in the strategy area. Before the strategic plan is developed in an existing organization, it may be important to assess the current reward system and to determine what kind of behaviors, climate, and structure the system supports. This step is needed so that when the strategic plan is developed it is based on a realistic assessment of the current condition of the organization and the changes that are likely to be needed to implement the new strategy. This point is particularly pertinent to organizations that are considering going into new lines of business, developing new markets, and acquiring new divisions. Often, new lines of business require a different behavior and therefore a different reward system. Simply putting the old reward system in place in the new business is often not good enough and indeed, can lead to failure in the new business. On the other hand, developing a new reward system for one part of an organization can cause problems in other parts because of the type of comparisons which are made between the different parts.

4

At this point it is impossible to go beyond the conceptual level in discussing reward systems and strategy. There are some useful typologies or menus available. These typically identify different strategies and then match them to reward system practices (see e.g., Galbraith & Kazanjian, 1986; Lawler, 1990). These are useful from an organization design point of view but are largely untested. Until more research is done in this area it is impossible to make any definitive statements. This is likely to be an increasingly important area for practitioners if not for researchers. The field of strategy is developing and companies are increasingly focusing on how they can gain a competitive advantage (Porter, 1990). There is good reason to believe that human resource management practices in general and reward systems in particular are central to the implementation and effective operationalization of strategies. Thus it is important to determine how reward system practices fit with different business strategies.

Reward Systems and Organization Design The literature on organization design increasingly uses systems theory and fit models which stress the importance of the congruence among different subsystems of the organization (see e.g., Galbraith, 1973; 1977; Nadler & Tushman, 1988;). Most of these approaches in turn identify the reward system as one of the major systems in an organization that needs to be driven by the strategy and aligned with the other systems. The other systems mentioned usually include the information system, the decision structure, the approach to work design, and the human resource management system. The literature usually goes on to argue that it is the degree of alignment that determines how effective an organization actually is. This of course raises the issue of what constitutes good alignment not to mention the issue of what data exist to support this assertion. At this point there is little research evidence to support the congruence or fit argument. Nevertheless it is an attractive notion and one that has high face validity particularly if system thinking is

5

accepted. It is easy, for example, to argue that an approach to organizing which centers all decisionmaking at the top may need a different reward system than an approach based on pushing decisions to lower levels (Lawler, 1990). In addition it is easy to accept the argument that the organization probably also needs a different approach to information and control. A substantial literature does exist that focuses on the relationship between reward systems and the degree to which participative management is practiced (see e.g., Lawler, 1981, 1990, 1992). A brief examination of the literature will serve to help focus on some of the organization design issue that arise with reward systems. This literature argues that if an organization wishes to operate in a participative manner it needs to change more than just its decision structure and information system flow. It needs to change all its systems including its reward system. This line of reasoning goes back to the early writings on the Scanlon plan (e.g., Lesieur, 1958) and the writings of McGregor (1960). They argue rather convincingly that for participative management to be practiced effectively a different approach to pay for performance is required. They go on to suggest that the correct approach is to pay bonuses based upon group or plant wide performance. The argument for this essentially rests upon the point that traditional pay plans support individual excellence at the cost of team performance and that for participation to work, team performance needs to be rewarded. The evidence on the impact of gainsharing plans (e.g., Graham-Moore & Ross, 1983, 1990) tends to support this point and it will be reviewed later in this chapter. At this point it is cited merely to establish the link between reward system design and organizational design. The more recent writings on participative management have gone considerably beyond the argument that it requires a different approach to paying for performance. It argues that all the major features of a reward system need to be changed (e.g., Lawler, 1990). For example, it is argued that pay needs to be based more on the skills of the individual than on the jobs that they do, that fringe benefit

6

programs should allow choice, and that pay information should be public. Later each of these practices will be looked at separately, the important point to remember, however, is that part of the argument in favor of them rests on systems thinking; thus, to a degree they cannot be evaluated without studying them in their proper context. There is little direct evidence supporting the argument that reward systems need to be structured differently for participative management to be effective. There is some evidence about companies' actual reward system practices that appears to be supportive, however. It tends to show that participative management is being more widely practiced and that reward systems are changing in ways that are consistent with participative management (Lawler, Ledford & Mohrman, 1989; Lawler, Mohrman, & Ledford, 1992; O'Dell, 1987). For example, gainsharing has become increasingly popular in the last decade and skill based pay plans have replaced job based plans in many manufacturing locations practicing participative management (Gupta, Jenkins, & Curington, 1986; Lawler, Mohrman & Ledford, 1992). In summary, there is a substantial literature that focuses on the organizational systems issues that affect the impact of reward system practices. The problem with it is that it lacks a substantial empirical basis. Some research has looked at the effect of certain reward system practices on individuals and organizations. This is important work and will be reviewed later, but it is not entirely satisfying because it provides limited information on how the impact of reward systems is affected by the organizational context.

Impact of Pay System The first step in a detailed discussion of the impact of reward systems is to consider what behaviors they can affect in organizations. The theoretical research and work on reward systems has

7

focused on six factors influenced by reward systems which in turn influence organizational effectiveness. Attraction and Retention Research on job choice, career choice and turnover clearly shows that the kind and level of rewards an organization offers influence who is attracted to work for an organization and who will continue to work for it (see e.g., Mobley, 1982; Mobley, Hand, Meglino & Griffeth, 1979). Overall, those organizations that give the most rewards tend to attract and retain the most people (Lawler, 1971). This seems to occur because high reward levels lead to high satisfaction, which in turn leads to lower turnover and more job applicants. Individuals who are satisfied with their jobs expect to continue to be satisfied and, as a result, want to stay with the same organization. This interpretation is generally consistent with the work on equity theory as well as the work on expectancy theory and goal-setting (Adams, 1965; Lawler, 1973; Locke & Latham, 1990). The best performers represent a particularly interesting retention problem. To retain them a reward system must distribute rewards in a way that will lead them to feel equitably treated when they compare their rewards with those received by individuals performing similar jobs at a similar level of performance in other organizations. The emphasis here is on external comparisons because turnover means leaving an organization for a better situation elsewhere. One way to accomplish this is to reward everyone at a level that is above the reward levels in other organizations. This strategy has two drawbacks, however. First, in the case of some rewards, such as money, it can be very costly. Second, it can also cause feelings of intraorganizational inequity; better performers are likely to feel inequitably treated when they are rewarded at the same level as poor performers in the same organization, even though they are fairly treated in terms of external comparisons. Faced with this situation, the better performers may not quit, but they are likely to be dissatisfied, complain, look for internal transfers, and mistrust the organization.

8

The answer may lie in basing rewards on performance. This should cause the better performers to be satisfied and to stay with the organization. It is important to note, however, that to be satisfied not only must the better performers receive more rewards than poor performers, they must receive significantly more rewards because, as equity theory points out, they are likely to feel they deserve more. Research has shown that absenteeism and pay satisfaction are related (e.g., Baumgartel & Sobol, 1959; see Steers & Rhodes, 1978, for a full review), although the relationship is not as strong as the one between pay satisfaction and turnover. When the workplace is satisfying and high paying, individuals come to work regularly; when it is not, they do not. Several studies have shown that absenteeism can be reduced by tying pay bonuses and other rewards to attendance (Lawler, 1981; Pedalino & Gamboa, 1974; Schlotzhauer & Rosse, 1985). This approach can be costly, but is sometimes less costly than absenteeism. It is a particularly useful strategy in situations where both the work content and the working conditions are poor and do not lend themselves to meaningful improvements. In situations where work content or conditions can be improved, such improvements are often the most effective and cost efficient way to deal with absenteeism (Hackman & Oldham, 1980). Reward system policies are only one of several ways to influence absenteeism, but they are potentially effective if an organization is willing to tie important rewards with coming to work. Often this is easier to do than tying rewards to performance, because attendance is an easily measurable and visible behavior. Motivation of Performance When certain specifiable conditions exist, reward systems have been shown to motivate performance (Blinder, 1990; Lawler, 1971; Nalbantain, 1987; Vroom, 1964). What are those conditions? They are rather clearly articulated in the extensive literature on expectancy theory: important rewards must be perceived to be tied in a timely fashion to effective performance. In essence the argument is that

9

organizations get the kind of behavior that leads to the rewards their employees value (Kerr, 1975). This occurs because people have their own needs and mental maps of what the world is like. They use these maps to choose those behaviors that lead to outcomes that satisfy their needs. People are inherently neither motivated nor unmotivated to perform effectively; performance motivation depends on the situation, how it is perceived, and the needs of people. Expectancy theory consists of a series of fairly straightforward observations about behavior. Three concepts serve as the key building blocks of the theory. Performance-Outcome Expectancy. Every behavior has associated with it, in an individual's mind, certain outcomes (rewards or punishments). In other words, individuals believe or expect that if they behave in a certain way, they will get certain things. For example, individuals may have an expectancy that if they produce ten units, they will receive their normal hourly rate, while if they produce fifteen units, they will receive their hourly pay rate plus a bonus. Similarly, individuals may believe that certain levels of performance will lead to approval or disapproval from members of their work group or their supervisor. Each performance level can be seen as leading to a number of different kinds of outcomes. Outcome Attractiveness. Outcomes are differentially attractive to different individuals. This is true because outcome values result from individual needs and perceptions, which differ because they reflect other factors in an individual's life. Some individuals may value an opportunity for promotion or advancement because of their needs for achievement or power, while others may not want to be promoted and leave their current work group because of needs for affiliation with others. Similarly, a fringe benefit, such as a pension plan, may have great value for older workers but little value for younger employees on their first job. Effort-Performance Expectancy. Each behavior also has associated with it a certain expectancy

10

or probability of success. This expectancy represents the individual's perception of how hard it will be to achieve the behavior and the probability of his or her successful achievement of that behavior (Locke & Latham, 1990). For example, employees may have a strong expectancy (e.g., ninety-ten) that if they put forth the effort, they can produce ten units per hour, but that they only have a fifty-fifty chance of producing fifteen units per hour if they try. This expectancy is based upon the characteristics of the task and the situation, as well as such individual differences factors as an individual's sense of efficacy (Bandura, 1986). Putting these concepts together, it is possible to make a basic statement about motivation. In general, an individual's motivation to behave in a certain way is greatest when: 1.

the individual believes that the behavior will lead to certain outcomes (performance-outcome expectancy);

2.

the individual feels that these outcomes are attractive; and

3.

the individual believes that performance at a desired level is possible (effort-performance expectancy).

Given a number of alternative levels of behavior (ten, fifteen, or twenty units of production per hour, for example), an individual will choose the level of performance that has the greatest motivational force associated with it, as indicated by a combination of the relevant expectancies, outcomes, and values. This combining process is theoretically defined as the multiplicative combination of these components; if the value of any one component is zero there will be no motivation. In other words, when faced with choices about behavior, an individual goes through a process of considering questions such as: "Can I perform at that level if I try?" "If I perform at that level, what will happen?" and "How do I feel about those things that will happen?" The individual then decides to behave in a way that seems to have the best chance of producing positive, desired outcomes.

11

The expectancy model also suggests that satisfaction is best thought of as a result of performance rather than as a cause of it (Locke & Latham, 1990; Porter & Lawler, 1968). Strictly speaking, satisfaction does influence motivation in some ways. For instance, when it is perceived to come about as a result of performance, it can increase motivation because it strengthens people's beliefs about the consequences of performance. Also, it can lead to a decrease in the importance of certain outcomes, and as a result, decrease the motivation for those performances which are seen to lead to whatever reward becomes less important. In many ways, the expectancy model is a deceptively simple statement of the conditions that must exist if rewards are to motivate performance. It is deceptive in the sense that it suggests all an organization has to do is relate pay and other frequently valued rewards to obtainable levels of performance. Not only is this not the only thing an organization has to do, as is obvious from the research on pay for performance systems, it can be difficult to accomplish. In order for employees to believe that a pay for performance relationship exists, the connection between performance and rewards must be visible, and a climate of trust and credibility must exist in the organization. The reason why visibility is necessary should be obvious; the importance of trust may be less so. The belief that performance will lead to rewards is essentially a prediction about the future. For individuals to make this kind of prediction they have to trust the system that is promising the rewards. Unfortunately, it is not always clear how a climate of trust in the reward system can be established. As will be discussed later, some research suggests that a high level of openness and the use of participation can contribute to trust in the pay system. Motivation of Self Development Just as reward systems motivate performance they can motivate the learning of skills and the development of knowledge (Lawler, 1990). The key here is the same as it is with performance

12

motivation. Individuals need to see a connection between learning specific skills and a valued reward. Pay for performance systems may motivate learning and development because individuals perceive that they must develop their skills in order to perform effectively. If individuals feel they already have the skills, however, then a pay for performance system is unlikely to have this impact. Sometimes pay for performance systems may discourage individuals from learning new skills or motivate them to learn the wrong skills. This can happen when the skills that should be learned are not directly related to present performance and as a result are not likely to lead to a reward and indeed may decrease the chances of receiving a performance-based reward. The reward systems in most organizations are extremely hierarchial. In major U.S. corporations for example executives often make more than 100 times as much as lower paid employees (Crystal, 1991). Strong motivation is thus created to learn those skills that are perceived to lead to promotion. To counter this tendency some organizations are using skill based pay when they want individuals to add new skills that do not involve promotions. They also use them when they want employees to develop a broader understanding of how the organization operates (Jenkins & Gupta, 1985; Lawler & Ledford, 1985).

Culture Reward systems are one organizational feature that contributes to the overall culture and climate of the organization (Whyte, 1955). Depending upon how reward systems are developed, administered, and managed, they may cause the culture of an organization to vary quite widely. For example, they may influence the degree to which the organization is seen as having a human resources oriented culture, an entrepreneurial culture, an innovative culture, a competence based culture, and/or a participative culture. Reward systems have the ability to shape culture precisely because of their important influence

13

on motivation, satisfaction, and membership (Kerr & Slocum, 1987). The behaviors they elicit become the dominant patterns of behavior in the organization and lead to perceptions and beliefs about what an organization stands for, believes in, and values. Perhaps the most obvious link between pay system practice and culture concerns the practice of performance-based pay. The absence or presence of this policy as well as how well it is implemented can have a dramatic impact on the culture of an organization because it so clearly communicates to organization members the expectations of the organization about performance. Many other features of the reward system may also influence culture. For example, having relatively high pay levels may produce a culture in which people feel they are an elite group working for a top-flight organization. Introducing such innovative pay practices as flexible benefits may produce a culture of innovativeness. Finally, having employees participate in pay decisions may produce a participative culture in which employees are generally involved in business decisions and as a result are committed to the organization and its success (Lawler, 1981). Reinforce and Define Structure The reward system of an organization can help reinforce and define the organization's structure (Lawler, 1981, 1990). Often this feature of reward systems is not fully considered in their design. As a result, the reward system's impact on the structure of an organization is unintentional. This does not mean, however, that its impact on organizational structure is necessarily minimal. Indeed, the reward system will help define the status hierarchy and the degree to which people cooperate with people from other departments and with people from within their work area. It will also strongly influence the kind of decision structure that exists. The decision to cover a group of people with a pay for performance system is an important structural decision that causes both integration and differentiation (Lawrence & Lorsch, 1967). It can

14

have a particularly strong impact on the degree to which people feel they share a common fate with others in the organization. Alternatively, it can set them off or differentiate them from others, creating a kind of structural boundary that reduces cooperation and sharing with individuals and groups outside the boundaries of the pay system. If the system pays all employees the same based on a measure of their collective performance, it can integrate them and cause them to cooperate and work as a team. On the other hand if the reward system asks individuals to vie among themselves for a fixed amount of money that has been allocated for raises or bonuses it can serve to differentiate them from each other and cause them to compete. Cost Reward systems are often a significant cost factor in organizations. Indeed, the pay system alone may represent over 50% of an organization's operating cost. Pay systems involve direct pay and benefit costs as well as the costs associated with managing and operating the system. In the case of some pay systems, these operating costs can be quite high. Thus, it is important in considering a reward system's impact to focus on how high these costs are. Recently economists have argued that it is important to consider how costs vary as a function of the organization's ability to pay (Weitzman, 1984). For example, a reasonable outcome of a well designed pay system might be an increased cost when the organization has the money to spend and a decreased cost when the organization does not have the money. This can help an organization avoid layoffs and provide individuals with greater job security. An additional objective for an organization might be to have lower overall reward system costs than its business competitors.

Establishing Base Pay Most pay systems establish the base pay of individuals by focusing on the job they hold (Rock &

15

Berger, 1991). Job evaluation is used to establish a basis for the compensation of different jobs. This approach will be discussed in the next section as will an alternative approach that is based on the skills individuals possess. Job Evaluation Job evaluation is a formal procedure for ordering a set of jobs or positions in an organization with respect to their inherent value or worth in achieving the organization's objectives. This procedure is usually then linked to a rate of pay prescribed for that job. While the origins of job evaluation are variously attributed to the U.S. Civil Service Commission in 1871 (Patton, Littlefield, & Self, 1964) or to the work of Frederick W. Taylor in the 1880's (Pasquale, 1969), the earliest significant applications were found in the late 1930's with the growth of the labor movement (Patten, 1987). Its use in the private sector became widespread during World War II when the National War Labor Board permitted wage increases only to correct demonstrated "inequities" in wage structure (Livy, 1975). At about the same time most government jobs were being classified using some form of job evaluation system (Suskin, 1977). Generally, by the mid 1970's about three-fourths of U.S. firms were using some form of job evaluation (Bureau of National Affairs, 1976). Regardless of the specific type of job evaluation procedure used, almost all share a similar methodology. First, the jobs to be evaluated are carefully described using a thorough job analysis. Second, based on the description of the job, each job is evaluated with respect to its value or "worth" to the organization, and the jobs arranged hierarchically with respect to this evaluation. Finally, these results are used to set wage rates within the organization. The remainder of this section will focus on the second step in the process since it is the area which has been the subject of the most compensation research. Methods of Job Evaluation

16

There are four major methods of job evaluation: job ranking systems, job classification systems, factor comparison systems, and the point system. The first two of these methods are generally referred to as qualitative techniques, the second two as quantitative techniques. Job Ranking. Job ranking systems are the crudest of the job evaluation techniques. Based on their relative worth to the organizations, jobs are ranked as a whole from the highest to the lowest. Because the rankings are based on job descriptions, not specific job facts, they are highly subjective. This procedure can be difficult to implement in a large organization where many jobs are to be evaluated.

Job Classification. This approach involves developing a set of job grades or classes and then fitting the jobs in the organization into these classes. First, comprehensive grade descriptions are written. These descriptions must be written generally so that many different types of jobs can be classified. Evaluators then examine job descriptions as a whole and match them to the appropriate grade description. Two difficulties with this approach are immediately obvious. The level of generality at which grade descriptions are written may encourage individuals to exaggerate and aggrandize their duties and responsibilities to obtain higher classifications and hence higher base pay. Second, most jobs will have some duties that are contained in one grade description and some in another. The evaluator is then forced to average or to pick the closest match to the job. Factor Comparison Approaches. Factor comparison job evaluations, compare jobs with respect to job dimensions or job requirements which the organization has deemed deserve compensation. The skeleton of the system is formed by selecting "key" or "benchmark" jobs that have relatively stable content, are found in many organizations, and for which a prevailing wage is known. Some systems use the current wage rate for the benchmark jobs rather than the prevailing wage. These jobs are then ranked along each compensable factor and, based on these rankings, the prevailing wage is apportioned across

17

the compensable factors. Once the skeleton of the system is assembled, other jobs in the organization are compared factor by factor to the key jobs and given the dollar value thought to be appropriate for that factor. The dollar values assigned for each factor are then summed to reflect the worth of the non-key jobs. While cumbersome to develop and explain, this approach is relatively easy to administer as long as prevailing wage rates are constant. There are two serious weaknesses in this approach, however. First, wage rates for the key jobs must be equitably aligned both internally and externally to produce useful assessments of relative worth. Second, the apportionment of the wage of the key jobs across factors must be valid since they constitute the foundation for the entire system. Point Systems. The point system is the most widely used of the job evaluation approaches (Lawler, 1990). Like the factor comparison approach, compensable factors are first determined and given weights or points reflecting the relative importance of each factor in determining job worth. The factors are then divided into degrees, each accompanied by a detailed description of the job content and requirements, and the points associated with that factor allocated across them. The upper end of each factor scale represents the maximum points associated with that factor, the lower end either zero or some "base." Using job descriptions and/or task inventories, job evaluators then assess each job with respect to how much of each factor the job contains, match that assessment to the degree descriptions, and award the number of points associated with that degree. The relative worth of the job is then reflected in the sum of the points awarded for each factor. Criticisms of Job Evaluation Systems While job evaluation approaches are probably the most prevalent methods of establishing base pay, their usefulness, like the usefulness of any management system, depends on how well they achieve their objectives (Lawler, 1990). They have been increasingly subject to criticism. For example, Belcher

18

(1969) characterized them as based on "untested assumptions" and Mahoney (1991) described them as increasingly anachronistic. The criticisms discussed next address three different, but related areas: the reliability and validity of job evaluation techniques, the unintended consequences likely to accompany them, and the congruence between these techniques and other organizational systems. Psychometric Properties. The reliability of job evaluation instruments has been challenged because the evaluation process is inherently subjective (Beatty & Beatty, 1984). Investigations of the reliability of job evaluations appear in two clusters (Welbourne & Gomez-Mejia, 1991): research in the late 1940's when job evaluation systems were achieving prominence as compensation tools (e.g., Ash,1948; Chesler, 1948; Jones, 1948; Lawshe & Farbro, 1949; Lawshe & Wilson, 1947); and research in the 1980's as questions of equal pay for work of comparable worth were raised (e.g., reviews by Arvey, 1986; Madigan, 1985; Milkovich & Broderick, 1982; Schwab, 1980; Snelgar, 1983). The earlier studies generally report high reliabilities for the total scores and somewhat lower reliabilities for individual factor ratings. In addition, systems using few factors tend to have higher reliabilities than those using many factors. The later studies are more likely to report mixed results (Arvey, 1986). For example, Gomez-Mejia, Page, and Tornow (1982), examine seven different job evaluation methods and report reliabilities of .44 to .80. With respect to the validity of job evaluation techniques for producing assessments of the organizational value of jobs, Arvey (1986) characterizes the evidence as "sparse." Generally, the validation studies that do exist approach the validation process from either a convergent or a predictive validity perspective. In the first approach, a set of jobs are subjected to varying job evaluation techniques and the results compared (e.g., Gomez-Mejia, Page, & Tornow, 1982; Madigan & Hoover, 1986). In the second approach, job evaluation results are used to predict either the organization's existing wage structure or the prevailing wage rate as defined by wage and salary surveys (e.g., Dertien, 1981;

19

Schwab & Grams, 1985). Even if the results of both approaches are consistent and supportive (which they are not), this does not establish the construct validity of job evaluation as a measure of relative worth (Gupta & Jenkins, 1991). The first approach which looks at multiple job evaluation approaches simply demonstrates that different job evaluation techniques are measuring the same thing. Whether that "thing" is the value of the job to the organization remains to be seen. It may simply reflect the shared stereotypes, biases, training, or experience of the evaluators (Gupta & Jenkins, 1991). Rynes and Milkovich (1986) have argued that the "going rate" for jobs does not exist. Rather, it is the sum of a series of subjective judgments on the part of those who conduct and use salary surveys. Similarly, the hierarchy of wage rates within a firm is a product of history, negotiation, stereotypes, convenience, and rationalization. Thus the predictive approach to validation cannot prove that a job evaluation technique actually captures the "true" value a job adds toward meeting the organization's objectives. All it establishes is that the job evaluation approach reflects the subjective judgments and biases that have been institutionalized either in the marketplace or internally in the organization. In summary, the evidence to date leaves the construct validity of job evaluations in question. Reliability evidence is abundant, but the convergent and predictive validity data are less convincing. They are not sufficient to establish that the "true" worth or value of a job has been captured (Gupta & Jenkins, 1991). Unintended Consequences. A job evaluation system is a control system. Control systems are designed to influence and direct employee behavior (Lawler & Rhode, 1976). The behaviors they influence and direct, however, are not necessarily those that are intended or those that are functional for the organization (Kerr, 1975). Job evaluation systems are no different in this respect. Lawler (1986c) has highlighted a number of unintended and organizationally undesirable employee behaviors that are

20

often associated with the use of job evaluation systems. First, because job evaluation systems tie pay to those activities included in the employee's responsibilities, they necessarily highlight those activities that are outside that purview. "That's not my job" and "I'm not paid to do that" are often heard refrains for not doing things that further the organization's objective. Second, because the intent of job evaluation is establishing internal equity, these systems direct employee attention to the internal relationships among jobs. Rather than paying attention to the competition in the marketplace, employees concentrate on how the pay for their jobs differs from the pay of others and how they can improve their relative pay positions. Third, employees quickly realize that creatively written and aggrandized job descriptions and exaggerated task inventories can sometimes lead to pay increases more readily than actual changes in job duties. Thus they produce inflated descriptions which can raise their pay and the organization's costs. Fourth, job evaluation systems implicitly or explicitly assign a heavy weight to responsibility, reinforcing the idea of a management hierarchy. Promotion, therefore, becomes the surest way to increase one's pay. In many modern organizations that require technical excellence or a broad-based understanding of the organization, the "linear career orientation" engendered by job evaluation approaches works at cross-purposes with organizational objectives (Fuller, 1972). The development of career ladders and maturity curves for scientists as alternative base pay structures for some groups of employees attests to the failure of traditional job evaluation methods to reward employees for electing careers paths where they could only look forward to static or potentially declining levels of compensation. System Congruence. Mahoney (1991) argues that the notion of work is undergoing a fundamental shift. Viewing work as a set of standardized, programmed tasks, appropriate to the

21

industrialization of the early twentieth century, is inappropriate in a time of internationalized markets, rapid technological and product developments, and emphases on quality and timeliness. The demands of this new environment require new organizational systems for designing work, authority structures, and organizations. The compensation system must be congruent with these new systems. Job evaluation systems that reinforce power relationships and "pecking orders" based on pay may impede the success of organizations that need technical knowledge and innovation often found at the bottom of the organizations (Lawler, 1986c). Similarly, successfully managing organizations that require flexibility and adaptability from their employees to meet changing environments is at odds with job evaluation systems based on prescriptive job descriptions. Job evaluation systems that reinforce organizational hierarchies also make the use of participative management systems more difficult (Lawler, 1990). Employee participation has an inherent dimension of equality. To the extent that the job evaluation system forces differences in pay and, by implication, worth, the success of a participative system is jeopardized. Organizations are adopting team-based approaches to organizing to meet the new organizational realities (Lawler, Mohrman, & Ledford, 1992). The unit of management in these organizations is the team. Tasks, duties, and responsibilities are assigned to a team, not an individual. While they are performed by individuals, individual responsibility shifts among team members as internal needs and external demands shift. Traditional job evaluation, with its focus on the job performed by an individual, is fundamentally incompatible with this organizational design. In theory, the logic of job evaluation could be applied to a team as a whole rather than to the individual team member with all team members paid at the same rate. But this leads to the same problem discussed earlier with respect to job evaluation plans that focus on individuals: evaluating the relative importance of the contribution of one team versus another (Gupta & Jenkins, 1991). Whatever the strategy taken, the end result bears little resemblance to

22

the original notion of job evaluation as conceived by Taylor and others. Organizations using traditional job evaluation techniques leave little room for rewarding individual variations in behavior. The jobs for which individuals are paid are clearly specified, as are the ways the jobs are to be performed. Mahoney (1991) notes that job evaluation systems are appropriate where jobs are standardized and unvarying, and where differences in performance are difficult to attribute to individuals. In many organizations, however, differences in individual performance levels are large and observable. Rewarding these differences means focusing on the individual which is in fundamental opposition to the basic tenet of job evaluation systems, that of paying the job. Conclusion. There is always likely to be a role for job evaluation in the administration of compensation. It can provide an internally justifiable wage structure for organizations that are structured around relatively stable collections of tasks. As organizations restructure to meet the new demands of a rapidly changing environment, it is less likely to be used because it fails to recognize differences in individual capability and it reinforces a hierarchial approach to management. Skill-Based Compensation Systems The classic works of Katz and Kahn (1966) and Emery and Trist (1965) which apply the logic of open-systems theory and sociotechnical systems theory to organizations, focus attention on the way work is designed and arranged. They focus particularly on the role of groups and on the design of work for groups or teams that are given the responsibility for the production of a whole product (Lawler, 1978; Walton, 1980). In these self-managing teams, members are expected to function interdependently to accomplish this end. The rotation of individuals among tasks or jobs within a team is expected to be frequent, each team member taking on responsibilities as needed and as he or she is able. Thus, team members are expected to become flexible members of the group by becoming multiply-skilled in the tasks that fall within the purview of the group.

23

The self-managing team approach was first used in a number of new plants, but it is now widely used (Lawler, Mohrman, & Ledford, 1992). When combined with other management practices and a management style which stresses employee participation, employee growth and development, it can produce a high-involvement work culture (Lawler, 1986a, 1992). Given the underlying premises of the job evaluation approaches and the problems associated with their use, a different approach to establishing base pay is needed in organizations that use teams and high-involvement management. Skill-based pay or pay-for-knowledge compensation systems seem to be the preferred approach (Lawler, 1990). Rather than paying individuals for the specific job they are performing, they are paid at a rate based on the repertoire of jobs or tasks they can perform, that is, their knowledge and mastery of different jobs in the organization. Pay-for-knowledge or skill-based compensations systems can be grouped into two broad types: multiskill based or breadth systems, where pay levels are linked to the number of different skills a worker learns and can perform in the organization; and increased knowledge or depth based systems, where pay levels are linked to increased knowledge and skill within the same general job category. Under a multiskill-based pay system, a new employee is paid a starting rate when first hired. When the employee masters a skill or task, a pay increase is awarded. As new skills or tasks are mastered, additional pay increases are given until the individual masters all the tasks required in the work team or all the tasks in the plant. Organizations have tried a number of variations on this basic theme. Sometimes the pace at which employees can learn new skills is restricted. In some instances, the proficiency with which a mastered skill is performed is used as a basis for awarding the increment associated with learning that skill. Sometimes the increments associated with all tasks are the same, sometimes the increments vary with the complexity of the task or the average time required to learn the task. Whatever the particular

24

details, multi-skill based pay plans tie individual pay levels to the number of skills or tasks the individual employee can perform. Increased-knowledge based systems, on the other hand, base individual pay on the depth of knowledge in a particular skill area. As individuals specialize in one skill area and acquire greater knowledge or proficiency in that area, increases in pay are awarded. Technical ladders in R&D organizations and organizations with apprenticeship systems in the skilled trades represent familiar forms of these systems. Less familiar are the depth based systems in high involvement organizations where a worker chooses to specialize in a particular skill area, gaining expertise in that particular job and then acting as a trainer, a resource to others learning that skill, and as the technical expert when problems arise. While no experimental evidence on the impact of skill based approaches to base pay exists, a number of recent empirical case studies (Gupta, Jenkins, Curington, Clements, Doty, Schweitzer, & Teutsch, 1986; LeBlanc, 1990; Ledford, 1985; Ledford & Bergel, 1990; Ledford, Tyler, & Dixey, 1990; Tosi & Tosi, 1986) and theory pieces (Jenkins & Gupta, 1985; Ledford, 1991; Lawler & Ledford, 1985; Tosi & Tosi, 1986) suggest they can offer advantages over traditional job-evaluation approaches for both employers and employees . Advantages for the Organization. The single biggest advantage of skill-based pay systems in a production setting is increased flexibility (Jenkins & Gupta, 1985). Because workers know more than one job, they can move with ease from one task to another as the need arises. As market and product demands change, workers can move from one set of jobs to another. As employees are absent, their roles can be filled temporarily by other team members, reducing the need to overstaff for absenteeism, fluctuations of supply and demand, and day-to-day uncertainties. Because in skill-based pay systems, human resources track organizational needs more closely, organizations can enjoy leaner staffing and

25

reduced overall labor costs. Organizations using skill-based pay systems also report higher quality output (Gupta, Jenkins, & Curington, 1986). Because workers are multiskilled, they have a better idea of how the different tasks fit together, and thus they have a greater understanding of the importance of quality in their performance. In addition, because one of the skills frequently learned in multiskill systems is quality assurance, workers know the quality assurance requirements for the task and can distinguish between acceptable and unacceptable products. Furthermore, as Jenkins and Gupta (1985) note, because of the frequent movement and rotation among jobs, operators producing low-quality output today may find they produced tomorrow's headaches when they move to quality assurance or to rework and repair. Absenteeism and turnover tend to be lower in organizations using skill-based pay. For example, in a study of plants using skill-based pay systems (Gupta, et al., 1986), seventy-four percent of the companies in the sample believed that their absenteeism rates were lower than they would have been without the skill-based pay system and sixty-nine percent reported that their voluntary turnover rates were lower. Many employees are more satisfied under these systems because they have a greater degree of control over their own lives and rates of pay. Also, because employees have a greater understanding of the entire operation in their organization, their intrinsic satisfaction may increase and they are better able to see the importance of their attending work regularly. Turnover is decreased in part by the increased satisfaction experienced by employees in these organizations and in part because, in March and Simon's (1958) terms, the ease of movement has been decreased with the addition of firm-specific training (Becker, 1975; Oi, 1983). Finally, these systems often produce higher pay rates. As employees become more broadly knowledgeable about the operations of the organization by working in different parts of it, the potential for them to become more self-managing is enhanced. They are in a position to control their own behavior, cooperate with others in the organization, solve problems,

26

and participate in decisions about how the organization operates. The more self-managing employees become, the fewer levels of management that are needed and in fact, the fewer managers, supervisors, staff and support personnel that are needed. In the Gupta et al. (1986) study, sixty-one percent of the companies responding said they would need more first-line supervisors without a skill-based pay system, sixty-three percent said they would need more skilled trades employees and thirty-two percent said they would need more managers. Advantages for Employees. From the employees' perspectives, a number of advantages may be gained from skill-based pay systems. Perhaps the most noticeable advantage is an increase in satisfaction. Research has shown that employee satisfaction levels are generally higher in skill-based pay companies than they are in similar companies using traditional pay systems (Lawler, 1990; Ledford, 1991). Because employees in skill-based pay plan are generally paid more than they would be under traditional compensation systems, this may reflect the strong correlation between pay satisfaction and pay level (Lawler, 1971). It is also possible that the increase in satisfaction derives from employees feeling treated as individuals and being able to influence their pay. Increases in the intrinsic motivation to perform may result from the enhanced job enrichment inherent in skill-based pay systems (Jenkins & Gupta, 1985). As individuals rotate through different jobs in the organization, either learning new skills or exercising the increased flexibility in the skills they possess, there is a natural increase in job variety, autonomy, task identity, and task significance with a consequent increase in motivation (Hackman & Oldham, 1980). Another employee benefit of these systems is an increased feeling of self-worth and self-esteem. The organization is demonstrating its regard for employees by the investment of resources in them and by providing the opportunity to learn and to grow. As people's talents develop, they feel better about themselves.

27

The increased flexibility and leaner staffing skill-based pay plans promote may also create increased job security for the employees. The organization is able to maintain a more stable work force, redeploying people in times of low or shifting demand. Costs of Skill-Based Pay Systems. The advantages of skill-based pay systems noted above do not come without costs. First, the pay rates that these systems produce tend to be higher on average than they would be under a traditional pay system. By rewarding skill acquisition, these systems encourage employees to become more valuable to the organization by mastering new skills. As employees master new skills, their average hourly pay increases. Second, because the systems encourage employees to learn multiple skills, the organization must make a larger investment in employee training than is the case in traditional systems. This investment can be quite substantial, involving hiring and paying trainers, lost time from work during training, and costs of less than fully competent individuals performing tasks. To some extent the training costs are mitigated because training is often provided by members of the work team as a normal part of the job, but the performance losses are real, at least in the short-run. In addition, because the work force is in flux, additional burdens are placed on the human resource management system which must coordinate the movement of employees from job to job and insure a proper balance between production and employee skill acquisition. It is especially important that the planning system support the goals of the pay system. The human resource planning systems must encourage movement of employees across jobs, and emphasize job rotation. Otherwise, employees quickly become frustrated and distrustful when they are ready to master a new skill and the opportunity to do so is not available to them. Third, skill-based pay systems place extra demands on the performance appraisal system. Employees must be evaluated more frequently than in traditional systems; they must be evaluated for their ability to do multiple jobs; and at times they must be evaluated by multiple appraisers. Because of

28

the increased empowerment of employees, inadequacies in the performance appraisal system are likely to become obvious, and pressure brought by employees to rectify these inadequacies. Deficiencies in the performance appraisal system will also have the effect of certifying less than fully skilled individuals, producing "jacks of all trades" but "masters of none." Fourth, the administrative burden is greatly increased with skill-based pay systems. The pay rates of employees change much more frequently and less predictably in skill-based pay systems than in traditional systems. Records must be maintained to keep track of which employees can do which jobs so that full advantage of the system can be realized. Fifth, because the system abandons the traditional notion of "paying the job" and focuses on "paying the individual," market comparisons become more difficult. When individuals perform jobs only part of the time and in different combinations, it is often impossible to find benchmark jobs in other organizations in order to peg pay rates. Furthermore, as multi-skilled employees are worth more to the organization than are singly skilled employees, simple market comparisons are inappropriate on their face. As skill-based pay systems mature, another difficulty develops. When a skill-based pay system is young, there are many jobs for employees to learn. But as the system matures, the more ambitious employees may have mastered all available tasks, with no place to go. While "topping out" occurs in traditional pay systems when individuals reach the top of their pay grades, it generally occurs much later in an employee's career. This problem is exacerbated by the fact that in skill-based pay systems employees may learn to expect more frequent pay raises than in traditional pay systems. Ensuring that employees remain proficient in the skills they learn can also present problems for the organization. When organizations are faced with pressure to produce, employees may not be allowed to move from job to job as freely. Skills that are not used for extended periods of time may atrophy and

29

the organization thus may compensate individuals for skills they either no longer possess or cannot perform as effectively as when originally learned. Finally, when organizations introduce technological innovations or other changes that make previously acquired skills no longer valuable to the organization, they must fashion fair and equitable methods to adjust employees' wages without penalizing them for decisions taken by management. One approach is to freeze the employee's rate until new skills are acquired to replace the obsolete skills. Developing and Implementing a Skill-Based Pay System Setting up a skill-based pay system is not an easy task. There is very little institutionalization of these programs. Traditional job evaluation systems such as the point system or the Hay plan which are more or less the same from one organization to the next and share a well-understood technology for their development and installation. Each skill-based pay plan is specific to the organization that uses it. Jenkins and Gupta (1985) note that there are probably as many variations of skill-based pay systems as there are organizations using them. While no handbook exists specifying the "correct" way to design a skill-based pay system, Ledford (1990) and Lawler (1990) have enumerated a set issues potential adopters must address and design "steps" that make a successful plan more likely. The first issue to be considered is how the skill-based pay system will fit with the organization's culture, its other management and technological systems and its business objectives. Generally, skillbased pay systems are most effective when organizations have or are moving toward a participative organizational culture. The breadth of experience and skill that skill-based pay systems produce in employees provides them a perspective for organizational decision-making. This perspective enhances employee problem-solving skills and cooperation, an important advantage where technological interdependence is high and teams are used. Second, the organization must identify the skills that are required for it to operate effectively.

30

This involves a careful analysis of the tasks that need to be performed in the organization. These tasks are then grouped or clustered into sets of skills (often called skill blocks) for which the organization is willing to pay more money. This step is critical to the ultimate success of the plan; the structure of the skill blocks determines (a) how well the pay plan fits the organization's technology and (b) the skeletal structure that provides the basis for most of the plan's detail (e.g., pay levels, certifications, training modules, etc). Third, the organization must identify the optimal skill profiles for all employees in the organization. The organization must specify how many employees with what combination of skills are needed to accomplish the organization's business objects. In doing so, the organization must consider its technological constraints. For example, manufacturing cells which are self-contained units making a set of products for a specific market require employees who possess all the skills used within the cell, but not necessarily those in other cells. By contrast, workers in continuous process technologies (such as those used in food processing and pharmaceutical production) are highly interdependent and must respond quickly to problems that may appear anywhere in the process. Thus, the organization needs employees who possess skills to work throughout the facility. Fourth, the organization must develop a mechanism for pricing each skill and skill-block. In some cases, the skills can be identified and priced by going to the labor market and determining the hiring rate for individuals with specific skills. In others, the "value-added" by the skill can serve as a guide. Regardless of the specific details of pricing the skill, this is one issue that cannot be sidestepped because it will form the basis for both internal and external equity assessments on the part of the employees. Fifth, the organization must develop a set of rules about the sequence in which skills can be learned and the rate at which they can be learned by employees. Because employees' wages are

31

determined by the number of skills they possess, they are naturally concerned with what they must do to acquire additional skills. Having too many employees learning new skills at the same time necessarily means that the organization is operating at a substandard level. Having too few employees learning new skills reduces the flexibility that organizations want from these systems. These access rules also have implications for the kinds, amounts, and timing of the training needed to impart the skills. Sixth, the organization must develop methods and procedures to determine when and whether an employee has acquired a new skill. This is often the most contentious issue in skill-based pay systems. The criteria and methods used to determine skill assessment must be developed in light of the nature of the job that is under consideration. It is generally believed that a work sample test in which the employee performs the tasks relevant to the skill is the best approach to skill certification. On the other hand, some skills, particularly those involving cognitive skills (for example, reporting and logging procedures), may be assessed more effectively using paper and pencil tests. When the testing of employees or skill certification is conducted (on demand or at set intervals) and how long employees must wait to be retested if they fail the certification must also be determined ahead of time. Finally, the organization must make a determination of who conducts these skill assessments and certifications. In traditional pay systems, the responsibility for pay decisions usually lies with management. In organizations high-involvement organizations, however, the responsibility may reside with co-workers and the teams. Finally, the organization must develop policies and procedures for decertification, i.e., for situations when individuals will no longer be paid for skills they previously mastered. This could occur, for instance, if there are changes in the technology the organization uses to produce its goods. It could also occur because the organization's product or service becomes obsolete. It could occur when individuals move from one work area to another, eliminating the need for the individuals to have those

32

skills. Furthermore, some skills may atrophy or be lost if they are not used regularly. Since it is not in the organization's interests to pay for unneeded skills, procedures for decertification must constitute integral components of the pay system design. Overall, developing a pay system that pays employees for the skills they possess requires a considerable amount of thought and system development. If properly done, however, there is evidence to support the view that pay can be used to reward individuals for developing the skills that help the organization function effectively, while at the same time furthering employee growth, development, and satisfaction. Compared to the amount of research on job evaluation, however, relatively little research has been done on skill-based pay. It is clearly an area where more research is needed.

Pay For Individual Performance The idea of paying for performance is so widely accepted that almost every organization says that it practices it. A survey of 557 large U.S. corporations found that eighty percent of them rate pay for performance as a very important compensation objective (Peck, 1984). The major reason for the popularity of paying for performance is the belief that it can motivate job performance and increase organizational effectiveness. The research evidence clearly supports this view. There has been, and continues to be, considerable evidence that pay can be a particularly powerful incentive (Blinder, 1990; Jenkins, 1985; Lawler, 1971; Locke, Feren, McCaleb, Shaw, & Denny, 1980; Nalbantian, 1987). Studies show productivity increases of between 1% and 35% when pay for performance systems are put into place (Lawler, 1990). Although pay for performance is often treated as a single approach, there are, in fact, many different approaches to paying for performance (see e.g., Blinder, 1990; Jenkins & Gupta, 1982; Lawler, 1990) which have very different consequences. They can be easily classified based upon the level of performance on which they focus: individual,

33

organizational subunit, or total organization. Within each of these general approaches, there are literally hundreds of different approaches to relating pay to performance. The focus in looking at the different approaches to paying for performance will be on their impact on individual and organizational behavior. Pay for performance systems are not stand alone systems; their impact is partially determined by the organizational context in which they operate. Thus as different approaches to pay for performance are reviewed, consideration also will be given to how they fit with different management styles and organization designs. Paying for Individual Performance There are two common approaches to paying for individual performance. The first, incentive pay, is declining in popularity. The other, merit salary increases, remains very popular (see, for example, Lawler, Ledford & Mohrman, 1989; O'Dell, 1987). They are similar in that they base monetary rewards on individual performance. They are different in how they measure performance and in how they adjust an individual's pay according to performance. Table 1 summarizes the characteristics of the two approaches. Incentive Pay Incentive plans pay employees bonuses based on the number of units produced. They are perhaps the most direct way to relate pay to performance. There is a great deal of evidence, much decades old, that incentive pay can motivate individual behavior (see e.g., Lawler, 1971). There is also good reason to believe that incentive pay plans can attract and selectively retain good performers because they end up being paid more. The literature on pay incentive plans is full of detailed descriptions of the counterproductive behaviors which piece-rate incentive plans produce (see e.g., Whyte, 1955). Most of the earlier accounts are from the manufacturing world, but the same kind of issues arise when sales and service personnel are

34

put on incentive pay (e.g., Babchuck & Goode, 1951). In many respects, theses behaviors are caused not so much by the concept itself, but by the way it is managed. Nevertheless, it is difficult to separate the practical problems with particular plans from the general idea of incentive pay. A brief review of the major problems with incentive plans follows. Beating the Systems. Numerous studies have shown that when piece rate plans are put into place an adversarial relationship frequently develops between system designers and employees (Lawler, 1971). Employees engage in behaviors to get rates set so as to maximize their financial gains relative to the amount of work that they have to do. They work at slow rates to mislead the time study expert studying their jobs. They hide new work methods or new procedures so the job will not be restudied. In addition, group norms develop about productivity and employees set limits on their production. Those who go beyond this limit are socially ostracized and may even be physically punished (Whyte, 1955). Unfortunately for the organization, this limit often is set far below that people are capable of producing. Other forms of gaming include producing at extremely low levels when the rates are set at levels that the employees consider too difficult to reach and using union grievance procedures to eliminate rates that are too difficult (Whyte, 1955). Another version of gaming involves doing only what is measured. In the case of production workers, this may mean not doing clean up and material handling work. In the case of sales persons, it may mean not doing customer service activities and tying up customers so that other sales persons cannot get the sale (Babchuck & Goode, 1951). Divided Work Force. Since many support and nonproduction jobs do not lend themselves to incentive pay, a typical organization with incentive pay has part of the work force on it and part of the work force not on it. This often leads to a we/they split in the work force that can be counterproductive and lead to noncooperative work relationships (Lawler, 1990). This split is not a management-worker split, but a worker-worker split that horizontally divides the organization. It can lead to incentive

35

employees complaining about materials handling, maintenance, and other employees whom they depend on for support. This split can also influence the kind of career paths people choose. Often, individuals will bid for and stay on incentive jobs even though they do not fit their skills and interests. The higher pay of incentive jobs may additionally cause individuals to be inflexible when asked to change jobs temporarily and cause them to resist a new technology that calls for a rate change. Maintenance Cost. Because incentive plans by themselves are relatively complicated and need to be constantly updated, a significant number of people are needed to maintain them. Maintaining incentive systems is further complicated by the adversarial relationship that develops between workers and management. Since employees try to hide new work methods and avoid changes in their rates (unless, of course, it is to their advantage), management must be extremely vigilant in determining when new rates are needed. In addition, when a technological change is made or a product is introduced, new rates must be set. While no good estimates exist of how expensive it is to maintain an incentive pay plan, it is likely that the costs vary widely from situation to situation. Organizational Culture. The combined effects of dividing the work force into those who are and are not on incentive pay and the adversarial process of rate setting can create a hostile, differentiated organization culture (Whyte, 1955). In particular, they produce a culture of low trust, lack of information sharing, conflict between groups, minimal support for joint problem solving, and employee inflexibility because individuals want to protect their pay rates. Overall, incentive pay generally works against creating a climate of openness, trust, joint problem-solving, and commitment to organizational objectives. Small Group Incentive Plans Closely related to individual incentive plans but much less popular are small group incentive plans (Lawler, Mohrman, & Ledford, 1992). They differ in that small group plans tend to fit situations

36

where outcomes are based on the performance of small employee groups and individual performance is not easily measurable. They are usually less effective in motivating performance because pay is less directly related to the performance of individuals. They can, however, be quite effective if the groups are kept small (Lawler, 1971). In general, they suffer from all the same problems as the individual incentive plans do, because they too are based on a top down management process, engineered standards and adversarial relationships. Conclusion The decades of research on incentive pay clearly demonstrates its strengths and weaknesses. It makes it clear that the installation of incentive pay is, at best, a mixed blessing. Although incentive pay may improve productivity, the counterproductive behaviors, the maintenance costs, the division of the work force, and the poor culture it engenders, may make it a poor investment. Many organizations have dropped or decided not to implement this approach simply because the organizations have concluded the negative effects and maintenance costs outweigh the performance increases performance incentive systems typically produces. Incentive pay clearly fits some organizational situations better than others. It best fits situations where the work is designed for individuals or in some cases for small groups. It best fits work that is simple, repetitive, and easy to measure comprehensively. Incentive pay tends to fit situations where the nature of the work is stable, where it can be carefully studied, and there is rarely the need to revise standards and payment approaches. More than any other system, it differentiates the organization to create isolated individuals or small groups who often feel in competition with one another. Thus, it fits best where the need for integration is negligible or where other mechanisms can be used to produce it (Lawler, 1981). Finally, incentive pay has usually been associated with the control approach to management and

37

it may be this fact that it has produced so many problems. It is possible that if it is used with a more involvement oriented management style that stresses fairness, due process, and participation incentive pay might be more effective (Lawler, 1990). It clearly has the power to motivate performance when individuals believe a fair and clear relationship exists between pay and performance. Merit Salary Systems Merit pay is the most widely used approach to paying for performance (see e.g., Lawler, Ledford, & Mohrman 1989; O'Dell, 1987). Merit pay systems typically give salary increases to individuals based upon their supervisor's appraisal of their performance. Their purpose is to affect motivation and to retain the best performers by establishing a clear performance reward relationship. In light of their popularity, it is surprising that so few studies have examined their impact on performance experimentally. Of the 77 studies reviewed by Heneman (1990), only six looked at the performance impact of the installation or removal of a merit system. Despite the widespread adaption of merit pay, there is considerable evidence that in most organizations merit pay systems fail to create a clear relationship between pay and performance (see Heneman, 1984, 1990; Milkovich & Wigdor, 1991). As a result, they also fail to produce the positive motivational effects expected of them. A number of reasons explain why merit pay systems often fail to relate pay and performance. Poor Performance Measures. Fundamental to an effective merit pay system are credible, comprehensive measures of performance. Without these, it is impossible to relate pay to performance in a way that is motivating. There is considerable evidence that performance appraisal is not done well in most organizations and, as a result, good measures of individual performance do not exist (see e.g., Devries, Morrison, Shullman & Gerlach, 1981; Meyer, Kay & French, 1965; Milkovich & Wigdor, 1991; Mohrman, Resnick-West & Lawler, 1989). There is also substantial evidence that even if good measures of performance exist, they often are distorted, ignored, or misused to accomplish other ends (Longnecker,

38

Sims, & Gioia, 1987; Mohrman & Lawler, 1983; Murphy & Cleveland, 1991). In the absence of good objective measures of individual performance, most organizations rely on the judgments of managers. These judgments are often seen by subordinates as invalid, unfair and discriminatory (Mohrman, Resnick-West & Lawler, 1989). When pay is based on performance measures that are not trusted, little is done to create the perception that pay is based on performance (Lawler, 1981). Several studies have found that when pay is related to performance appraisal results the nature of the appraisal is changed for the worse (see e.g., Meyer, Kay & French, 1965). Particularly troublesome is the finding that pay discussions tend to inhibit the discussion of development and learning. One study by Prince and Lawler (1986), however, suggests that discussing salary during a performance appraisal has either no negative effects or small positive effects on how well the appraisal is done and how satisfied the participants are with the appraisal session. It too, however, concludes that discussing pay makes it difficult to discuss development and suggests a separate meeting to consider it. Poor Communication. The salaries of most individuals in organizations are kept secret (Lawler, 1981). In addition, some organizations keep many of their pay practices secret. For example, it is common for organizations to keep secret such things as the amount of salary increases and what the highest and lowest raises are. Thus, the typical employee is often in the position of being asked to accept as an article of faith that pay and performance are related. Given pay secrecy, it is simply impossible to determine if such a relationship exists. In situations of high trust, employees may accept an organization's statement that pay increases are based on merit. Trust, however, depends on the open exchange of information; with secrecy, it is not surprising that many employees are skeptical. Poor Delivery Systems. The actual policies and procedures which make up a merit pay system often lead to actions which do little to relate pay to performance (Lawler, 1990). In addition, the policies and procedures are often so complex that they do more to obfuscate than to clarify the relationship

39

between pay and performance. The typical merit salary increase is particularly poor at relating pay and performance, because it allows for only small changes in total pay to occur in a single year. All too often only a few percentage points separate the raises given good performers and those given poor performers. This is particularly likely in periods of low inflation because salary increase budgets are usually small. Employees view pay increases as motivating only if they are large enough to be perceived as meaningful (Lawler, 1981). How large, is large enough? The idea of a pay raise threshold has been of interest for over two decades (see Zedeck & Smith, 1968) and has drawn heavily on psychophysical concepts, most notably Weber's Law (Gescheider, 1976). Eleven studies have attempted to investigate the notion of a pay raise threshold (Bowen, Worley, & Lawler, in press; Champlin & Kopelman, 1991; Corbett & Potocko, 1969; Futrell & Schul, 1980; Futrell & Varadarajan, 1985; Heneman & Ellis, 1982; Hinrichs, 1969; Krefting & Mahoney, 1977; Rambo & Pinto, 1989; Varadarajan & Futrell, 1984; Zedeck & Smith, 1968). These studies have produced estimates of just noticeable or psychologically meaningful pay raises ranging from 1.8% to 11.5%. Peck (1984) reported that in 1982 merit increases ranged from 5.0% to 14.3% with a median increase of 8.8%. If the value of a just meaningful pay raise is estimated at the midpoint of the range of values in the empirical investigations it comes out to 6.7%. This suggests large numbers of employees are receiving merit increases that are not noticed or psychologically significant. While there have been no empirical investigations of the impact of less than noticeable pay raises on employee motivation, it is reasonable to assume that they do not enhance pay for performance perceptions. Furthermore, when organizational pay policies (for example, guide charts) dictate that pay increases be based on factors other than performance, differences in increases between employees at different performance levels may be further reduced, again weakening pay for performance perceptions. Salary increase systems further compound the problem of relating present pay levels to present

40

performance by making past "merit payments" part of the individual's base salary so that it becomes an annuity. Indeed, employees current salary levels seem to be viewed by supervisors and employees alike as entitlements. While it may be acceptable not to grant an increase when employees perform poorly, salary levels are rarely, if ever, reduced. Thus, an individual can be a poor performer for many years after having been a good performer and still be highly paid. The less senior good performer, on the other hand, has to perform well for a number of years in order to achieve a relatively high pay level. Furthermore, as Schwab and Olsen (1988) demonstrated in a Monte Carlo simulation, the unreliability of the performance measure weakens the relationship between actual pay and actual performance over time when the merit payments become part of the employee's base salary. This can have disastrous effects for retaining outstanding performers. Because they are unable to increase their pay quickly, they often find it best to look for jobs elsewhere. The annuity feature leads to one other problem, topping out. After a long period on a job, individuals often reach a point where they are at the top of the pay range for their job. The effect is to eliminate pay as a motivator because it cannot be adjusted to reflect performance. Poor Managerial Behavior. Managers do a number of things that adversely affect the perceived and actual connection between pay and performance. Perhaps the most serious is the failure to recommend widely different pay increases for their subordinates when large performance differences exist (Kopelman, Rovenpor & Cayer, 1991). Some managers are unwilling to recommend very large and very small pay changes, even when they are warranted (Heneman, 1984). One reason for this seems to be the unpleasant task of explaining why someone got a low raise. Conclusion The impact of tying the pay levels of individuals to performance is a complex one that has been the subject of considerable research. Table 2 highlights the advantages, disadvantages, and fit issues

41

involved in both incentive and merit pay. The existence in most organizations of any one of the common problems that plague the administration of merit pay programs, however, is usually enough to destroy the belief that pay is related to performance and as a result eliminate the motivational impact of merit pay. In reality, the merit pay systems of most organization typically suffer from all or most of these problems. As a result, the policy of merit fails to improve performance. The problems with merit pay raise the question of whether it should be used. A number of the leading figures in the Total Quality movement have concluded that it should not (see e.g., Deming, 1986; Mohrman, 1990). They list many criticisms, but perhaps the most telling is that merit pay focuses attention on individual performance when in fact poor performance is often the result of poor organizational systems and processes. There is evidence that merit pay can have a positive impact when it fits the situation and is properly administrated (see e.g., Heneman, 1984, Kopelman, Rovenpor & Coyer, 1991). Like incentive pay, it focuses on individuals and as a result, does little to integrate the members of the work force. Indeed, the typical approach of allocating a salary increase budget to be divided among a small group of employees clearly sets up a competition for the available money. This can be a serious problem if the organization needs employees to cooperate in order for it to be effective. The research on individual incentive systems suggests they should be approached with a great deal of caution. As shown in Table 2, they can be a positive motivator but they do not fit most situations. Thus, they are likely to be used less frequently in the future and ultimately may be applied only to individual sales jobs, repetitive clerical jobs, and simple repetitive manufacturing jobs.

42

Pay for Organizational Performance Bonus payments based on the performance of an organization and stock ownership are timehonored and potentially effective ways to improve organizational performance. Proponents argue that they can improve motivation, build a work culture in which people are committed to and care about the organization's effectiveness, and finally, adjust the labor cost of an organization to its ability to pay (Weitzman, 1984). There is no question that in some instances, organizations have been able to accomplish just these outcomes as a result of tying the pay levels of individuals to organizational performance. It is far from simple, however, to design an effective approach. There are literally thousands of approaches to paying for organizational performance, and there are many complex organizational issues that must be dealt with if any plan is to be successful (Lawler, 1990). Historically, there have been three major approaches to paying for organizational performance. The oldest, profit sharing, is the approach of paying bonuses based on the profitability of the organization. This is undoubtedly the most widely accepted approach worldwide, and as will discussed later, has important advantages, as well as very important limitations. Closely related to profit sharing are stock ownership plans that grant individuals all or part of the ownership of the organizations for which they work. Employee ownership plans treat employees as investors, rewarding them when the organization does well in equity markets and reducing their wealth when the organization does poorly (Rosen & Young, 1991). Less common, but increasingly popular is gainsharing (Lawler, Ledford, & Mohrman 1989; Lawler, Mohrman, & Ledford, 1992; O'Dell, 1987). Gainsharing differs from profit sharing in two respects. First, it is combined with a participative approach to management, and second, it typically measures controllable costs or units of output, not profits or stock price in its approach to calculating a bonus (Lawler, 1990). Table 3 gives an overview of the characteristics of the three major organization

43

pay for performance systems. In the discussion which follows, consideration will first be given to what is know about gainsharing, then profit sharing and ownership plans will be considered. Gainsharing Gainsharing, a term originally coined by F.W. Taylor (Graham-Moore & Ross, 1990) has been used successfully for at least 50 years in hundreds of organizations (Bullock & Lawler, 1984; GAO, 1981; Graham-Moore & Ross, 1990; O'Dell, 1981, 1987). Both employees and companies have profited from gainsharing; companies in the form of reduced costs and employees in the form of bonus payments and improved job satisfaction. The original and best known gainsharing plan is the Scanlon Plan. Other gainsharing plans include Improshare and the Rucker Plan (see Graham-Moore & Ross, 1983, 1990, for detailed descriptions of these plans). In addition to these plans, many companies have their own custom designed gainsharing plans. In the typical gainsharing plan, a historical performance base level is established and is used as the basis for determining whether performance gains have occurred; hence the name "gainsharing." Typically, only costs controllable by employees are measured for the purpose of computing a performance gain. Any financial improvements in organizational performance above the historical "base" are then shared with all members of the organization. Unless a major organizational change takes place, the historical base stays the same during the plan; thus, performance is always compared to the time period before the initiation of the gainsharing plan. When performance is better than in the base period, a bonus pool is funded. When it falls short, no bonus pool contributions are made. In a typical plan, at least half of the bonus pool is paid out to the employees, while the rest is retained by the company. Payments are typically made on a monthly or quarterly basis with all employees receiving equal percentages of their base wages or salaries. There is no accurate estimate of how many gainsharing plans there are in the United States and

44

Europe. There probably are at least a thousand, and there is little doubt that their popularity has increased tremendously in the last ten years. One recent survey in the United States indicated that about 13% of all firms have them, and that over 70% were started in the last five years (O'Dell, 1987). Until ten years ago, gainsharing was used primarily in small manufacturing organizations. Much has been written in the United States about the success of older gainsharing programs in smaller companies such as Herman Miller, Lincoln Electric, and Donnelly Mirrors (Graham-Moore & Ross, 1983; Moore & Ross, 1978) which have been in effect over thirty years. During the 1970's, large companies such as General Electric, Motorola, Rockwell, TRW, Dana, and Firestone began installing gainsharing plan in some manufacturing plants. The trend of large corporations defining organizational units that have their own gainsharing is continuing, and is resulting in the adoption of more gainsharing plans (Lawler, Mohrman, & Ledford, 1992). The increased popularity of gainsharing is significant, and relates to an important feature of most gainsharing plans. They are more than just pay incentive plans; they are both a way of managing and an organizational development technology. They are a participative approach to management and are often used as a way to install participative management systems (Frost, Wakely, & Ruh, 1974; Lawler, 1986b, 1990). The Participative System. From the beginning, Joe Scanlon, the creator of the Scanlon Plan, emphasized that gainsharing fits a participative management style. In the absence of a change in employee behavior, there is no reason to expect a payout from the kind of formula which is typically developed in gainsharing plans. A payout requires an improvement in performance, and that improvement requires more effective behavior on the part of employees. Some improvement may be gained simply from the motivation that is tapped by tying pay to performance. This is particularly true in situations where the work is not highly skilled or interdependent and where effort is directly related to performance. In other situations, however, there are several reasons why a gainsharing plan without a

45

participative system is unlikely to produce an appreciable improvement in performance. First, the motivational impact of the plan may not be large because most gainsharing plans aggregate the performance of a number of people. As a result, the plan has only a small impact on the perceived relationship between individual performance and pay. The formula used is relevant here. Some plans use very simple formulas that focus on the relationship between labor input and productivity (e.g., Improshare); others use a more comprehensive set of cost measures (e.g., Rucker plan). Simple labor based plans are more likely to affect motivation directly because employees can see clearer relationship between their efforts and their bonuses. Second, in many cases, simple effort and good intentions are not sufficient to improve the operating results. Improvement in operating results requires a combination of people working harder, working more effectively together, sharing their ideas, and working smarter. This often requires a formal participative system that converts the motivation to improve performance into actual changes in the operating procedures of an organization. In the absence of new procedures or systems to accomplish theses changes, they rarely seem to occur. In traditional gainsharing plans such as the Scanlon Plan, the key to the participative system is a formal suggestion system with written suggestions and shop floor committees to review the suggestions (Frost, Wakley, & Ruh, 1974). Often, there is also a higher level review committee that evaluates those recommendations involving several parts of the organization and/or large expenditures. This system of committees is one way of assuring that new ideas will be seriously considered and, where appropriate, implemented. Recently, some organizations have combined gainsharing with self-managing teams and other participative management practices to produce an approach which is best called high involvement management (Lawler, 1986a, 1992). In this approach, employees make most of the operating decisions

46

and are rewarded for increases in their organization's effectiveness through the gainsharing plan. This approach gives employees a chance to influence the things that determine the operating results of the organization, a requisite condition for bonuses based on operating results to influence motivation. Research Results. The most important research result is that gainsharing plans usually work (GAO, 1981). The evidence suggests they are effective in about 70% of the cases (Bullock and Lawler, 1984). Lawler (1971, 1990) has summarized some of the common characteristics been found in studies of effective gainsharing plans: 1.

Coordination, teamwork, and sharing of knowledge are enhanced at lower levels.

2.

Social needs are recognized via participation and reinforcing group behavior.

3.

Attention is focused on cost savings, not just quantity of production.

4.

Acceptance of change due to technology, market, and new methods is greater because higher efficiency leads to bonuses.

5.

Attitudinal change occurs among workers, and they demand more efficient management, better planning, and good performance from their co-workers.

6.

Employees try to reduce overtime; to work smarter.

7.

Employees produce ideas as well as effort.

8.

When unions are present, more flexible administration of union management relations occurs.

9.

When unions support the plan, they are strengthened because better work situations and higher pay result.

10.

Unorganized locations tend to remain nonunion.

There also are things that gainsharing plans do not do as well as other approaches to paying for performance. Perhaps the most important is differentially attracting and retaining the best performers.

47

Because gainsharing plans do not pay better performers more, they do not necessarily motivate them to remain with the organization. Gainsharing plans do vary pay costs with the organization's ability to pay, but not as effectively as profit sharing since gainsharing plans can produce bonuses even when organizations are not profitable. Finally, gainsharing plans contribute to both integration and differentiation. They integrate the units they cover both vertically and horizontally since they treat all covered employees the same. On the other hand they tend to differentiate covered employees from the rest of the organization. Much is known about structuring gainsharing plans. A number of books and articles describe in detail how to develop formulas, how to introduce plans, and how to manage the process side of operating a gainsharing plan (e.g., Belcher, 1991; Graham-Moore & Ross, 1983; Moore & Ross, 1978). As a result, there is quite a bit of "how-to-do-it" knowledge. This is particularly true with respect to the Scanlon Plan. Indeed, a careful reading of the literature on this plan can make it possible for an organization to develop and install a plan without the help of a consultant. The research evidence also shows that certain situational factors favor gainsharing plans (Lawler, 1981). They include: 1.

Organization size. The plan is based on employees seeing a relationship between what they do and the pay they receive. As organizations get larger, this is harder to accomplish. Most successful gainsharing plans cover fewer than 500 employees. The plans also tend to cover operating units that can operate relatively independently of other organizational units.

2.

Performance measurement. In some organizations, good performance measure and a reasonable performance history simply do not exist and cannot be established. This is often true in organizations where rapid technological and market changes occur. For

48

these organizations, workable gainsharing plan formulas are difficult, if not impossible, to develop. 3.

Measurement complexity. Often performance can be measured only in very complex ways. The greater the complexity of performance measurement, the more difficult it is to make a plan work as no clear, easily understood connection between an individual's behavior and rewards can be established.

4.

Worker characteristics. Gainsharing depends on employees wanting to participate and wanting to earn more money. Most individuals have these goals, but not all do. Unless a substantial majority of the employees in an organization want the benefits the plan offers, it cannot succeed.

5.

Communication. For gainsharing to work, employees must understand the plan and trust it enough to believe that their pay will increase if they perform better. For this belief to exist, a great deal of open communication and education is needed. If an organization does not have these characteristics already, they must be established if the plan is to succeed.

6.

Management attitudes. Unless managers are favorably disposed to the idea of participation, gainsharing is unlikely to fit the management style of the organization. In some organization, the plan has been tried simply as a pay incentive plan without regard to management style, and it has failed because of a poor fit.

7.

Supervisory skills. Gainsharing requires supervisors to change. They are forced to deal with many employee generated suggestions and their technical competence is tested and questioned in new ways. Unless supervisors are prepared for and accept theses changes, the plan can fail. This point goes along with the general point that management must be

49

prepared to manage in a different way. As this list demonstrates, gainsharing does not fit every situation. Because manufacturing organizations are most likely to have these favorable conditions it is easy to see why, for so long, the installation of gainsharing plans was limited to these settings. Although much remains to be learned about how such plans should be installed in nonmanufacturing environments and in the public sector, it appears there are ways they can be designed to work in these settings (Graham-Moore & Ross, 1983; Schlesinger & Heskett, 1991). Indeed, it may be a more broadly applicable approach than has been generally assumed and demonstrated. Profit Sharing Profit sharing is better known, older and more widely practiced than gainsharing (Lawler, Ledford & Mohrman, 1989). In the United States, for example, data indicate that at least one third of all organizations have some form of profit sharing (O'Dell, 1987). Some definitions of profit sharing include it as a form of gainsharing; it is different, however, in two respects. It often does not have a participative management component and it does not use formulas that only include increases in employee controlled financial, or productivity related, performance. Given its popularity it is surprising that so little psychological research has been done on profitsharing. While much less popular, gainsharing has been much more extensively researched, probably reflecting the fact that gainsharing is associated with participative management. What evidence there is suggests that profit sharing plans typically are much less effective than gainsharing plans in influencing motivation and in producing the kind of social and cultural outcomes that were noted earlier as resulting from gainsharing plans (see e.g., Blinder, 1990). This is particularly true in large organizations where the line of sight and line of influence from individual performance to corporate profits is virtually nonexistent. As the size of the organization increases, the linkages of employee controllable performance

50

and employee rewards becomes so tenuous that there is likely to be little impact on individual motivation. In the typical profit sharing plan the line of influence problem is further compounded because most firms (estimates are about 85%) defer profit sharing bonuses by putting them into retirement plans (Blinder, 1990; Lawler, 1990; Metzger, 1964). Before dismissing profit sharing as totally useless from an organizational effectiveness point of view, it is important to note that there are three things even a deferred profit sharing plan in a large corporation can accomplish. First, there is the potential symbolic and communication value of paying employees based on organizational performance. Profit sharing plans effectively point out to employees that they are part of a larger organization and that cooperative effort is needed. Since corporate executives are often rewarded on the basis of organizational profits, it can also help to assure that there is an alignment between the rewards received by top management and those received by people throughout the organization (Foulkes, 1991). This can help avoid the counter productive vertical differentiation often created in organizations when executives receive large bonuses, while lower level employees receive little or none (Crystal, 1991). Second, some companies, most notably Hewlett-Packard, have effectively used their profit sharing plans as vehicles for educating employees about the financial performance of the business. When employees understand the financial performance of the business and are actually sharing in the profits, it brings alive for them the issues of what profits mean and how they are calculated and created. This in turn can increase their interest in learning about profits and organizational effectiveness and have an effect on the organization's culture. Third, perhaps the most important advantage profit sharing offers is that it makes the labor costs of an organization variable, and adjusts them to the organization's ability to pay (Weitzman, 1984). When profits go down, relative labor costs go down, and thus, rather than being fixed, labor costs

51

become, in part, variable. This is a particularly desirable feature for organizations that are in cyclical or seasonal businesses. In most western countries, changes in labor costs are handled through increases and decreases in the size of the workforce. When wages are high and fixed, there is no other way to control labor costs to reflect the company's ability to pay. With profit sharing, it is possible to reduce costs significantly without reducing the number of employees. Japanese firms have used this approach to for adjusting labor costs for decades. As is the case in Japan, it can allow an organization to make a much stronger commitment to employment stability and help it gain the advantages inherent in a stable workforce. Employee Ownership A number of plans exist that help get some or all of the ownership of a company into the hands of employees. These include stock option plans, stock purchase plans, and employee stock ownership plans (ESOP's). There is little question that stock ownership plans are increasingly popular and are being researched more. According to one study, some 11 million employees in over 8,000 businesses now own at least 15% of the companies employing them (Quarrey, Blasis, & Rosen, 1986; Rosen, Klein & Young, 1986). It is difficult to generalize about the impact of these plans because they vary widely in how much ownership employees receive and because the plan's impact is likely to depend on the organizational situation. Much of what has been said about the impact of profit sharing and gainsharing plans is relevant to the impact of ownership. In some situations there is reason to believe that ownership can have much the same impact as an effective gainsharing plan (Klein, 1987). In small organizations in which participative management is practiced it has a good chance of increasing organizational performance (Blinder 1990; O'Toole, 1979, Rosen, Klein & Young, 1986). The key to success is combining employee ownership with employee involvement since employee ownership plans per se typically produces a

52

weaker line of influence than does gainsharing (Blinder, 1990; Conte & Tannenbaum 1980). In a large organization with little employee ownership it may positively impact the structure by creating integration across the total organization if all employees are included in the ownership plan. Unlike profit sharing, it does not adjust costs to reflect the organization's ability to pay unless it includes an approach in which stockholders directly share in profits. It can help organizations raise capital and finance themselves. Indeed, most plans are probably installed because of the tax and financing advantages they offer (Blasi, 1988) rather than because of their potential for enhancing employee and organizational performance. Ownership can have a more positive impact on attraction and retention, however, than does profit sharing. In particular, if ownership is not easily transferable it can help to "lock" employees into the organization both financially and psychologically. Overall, based on the limited research available, there is reason to believe that ownership strategies can have a positive effect in a number of situations. Their usefulness, however, is likely to be situationally determined. For instance, in the case of small organizations they might make profit sharing and gainsharing unnecessary and if combined with an appropriate approach to employee involvement can contribute substantially to employee motivation. In a large organization they may contribute to a positive culture and to the integration of the organization. There is a need for considerable additional research on a number of issues including the impact of different approaches to structuring ownership and how the impact of ownership is affected by the characteristics of the organization. Theory and Predictions Existing theory and research can be combined to make some testable predictions about the impact of any pay for organizational performance system (Lawler, 1988b). As noted earlier, congruence theory assumes that an organization's effectiveness is related to the degree to which the different systems in an organization fit or are congruent with each other (see e.g., Galbraith, 1973). This calls for a

53

congruence among the way information, power, knowledge, and rewards are distributed in the organization (Lawler, 1986a). Organizational pay for performance plans move rewards for organizational performance downwards so that part of all employees' rewards depend on it. The implicit assumption is that this will change employee behavior in ways that will increase organizational performance. The major way it can change performance is by affecting employee motivation since it does little to reward training and developing new skills. It follows from the earlier discussion of motivation that organizational plans will increase organizational performance to the degree that they do the following: 1.

establish the belief that individual rewards are based on organizational performance;

2.

provide communications about organizational performance to all employees;

3.

establish ways for employees to influence organizational performance as it is measured by the reward system; and

4.

create opportunities for employees to learn how to contribute to organizational performance and how to interpret measures of performance.

These four conditions can best be viewed as combining multiplicatively and varying from zero to one. Viewing them as combining in this way is critical because it means that if any of them is completely absent, organizational performance will not improve as a result of installing gainsharing, profit sharing or stock ownership plans. As is frequently noted, these plans work only when individuals are motivated to perform differently. If one or more of these elements is missing, a pay change will not affect motivation because no linkage or line of sight and influence will exist between the reward and the inputs (i.e., effort, ideas) that employees in an organization can control. Working harder can improve individual performance but it may have no or a relatively small impact on organizational performance unless the organization already has good work methods and

54

systems. Often improvements in organizational performance require major system and method improvements; without them, increased individual employee effort may be wasted, resulting in increasingly frustrated employees. It is also important to note again that organizational pay for performance systems may not have a major impact on effort. Particularly in large organizations, they do little to increase the relationship between effort and rewards; thus, their impact on motivation is likely to be small or nonexistent. As the research shows, gainsharing, profit sharing, and employee ownership are more effective when combined with participative management (Blinder, 1990; Hammer, 1988). This follows from both motivation and congruence theory. In the absence of a participative management approach that gives employees a chance to increase organizational performance in ways other than working harder, it is difficult for them to see how they can influence organizational performance. Participative management can put into place two key elements: the ability to influence organizational performance in ways other than putting more effort into immediate job performance, and an understanding and trust of the relationship between improved organizational performance and a bonus or the value of an organization's stock. In the absence of these, it is unlikely that a perceived connection between improved individual performance and a pay reward will exist. One exception may be the case of very small organizations where a relationship between individual effort and organizational performance may be evident in the absence of a participative management style. In the case of small organizations, an organizational pay for performance plan might establish a line of influence and work as an incentive much like small group pay incentive plans do. Turning to the specifics of particular plans, the congruence argument suggests that the type of participative approach used needs to be matched to the type of pay plan used. Most gainsharing plans advocate suggestion programs of some type. Although suggestion approaches give employees some

55

additional power, they are very limited in the amount of power, knowledge, and information they move downward in the organization. They are, in essence, a parallel structure that does little to change the core operating style of the organization (Lawler, 1986a, 1988b, 1992). In this respect they differ appreciably from such approaches as work teams or semi-autonomous work groups that push a considerable amount of power and knowledge to the lowest levels of the organization (Hackman & Oldham, 1980). Teams can give employees a chance to influence a number of the decisions that directly influence organizational performance (Lawler, 1986a). The combined congruence-expectancy theory argument leads to some interesting predictions about which approach to participative management is needed to support different pay for organizational performance approaches. The Improshare formula is the simplest gainsharing formula since it looks only at hours of work relative to units produced. Because it is simple and straightforward, it is not difficult to establish a line of sight: employees know if they produce more product by either working smarter or harder they will earn more. Little extra communication and education are needed. Further, individuals do not need to influence many of the decisions which are made in organizations in order to reduce the amount of labor that goes into a product; they can simply work harder or smarter. In short, the Improshare plan has a chance of working as an incentive plan in the absence of a great deal of participation if there is a reasonable level of trust, the organization is relatively small, and the production process is relatively simple. Therefore, the use of quality circles or some other suggestion programs can provide a reasonable vehicle for reducing labor hours through better work methods. The situation is quite different for gainsharing formulas that look at multiple costs and/or involve complex computations and for profit sharing and ownership plans. The payments gainsharing plans produce are influenced by a wide range of factors often including such things as purchasing and pricing decisions. If employees are to influence their rewards, a level of participation that goes far beyond

56

quality circles and written suggestions is needed. They require at least the kind of participation that is present in self-managing teams and high involvement plants. Profit sharing plans and stock ownership plans are a simple extension of this argument. For them to be effective as motivators, a high level of employee involvement is essential. Other congruence issues come into play with respect to plan choice. The plan needs to fit the type of work the organization does. A labor only gainsharing formula, for example, fits only those operations where labor costs are dominant and can be reduced. Situations where other costs are important need more complex formulas that reflect other key operating costs. In the absence of a more complex formula there is the real danger that employees will focus only on reducing the measured costs and that unmeasured costs will increase. In summary, the congruence approach does not argue for a particular approach to relating pay to organizational performance. Instead it suggests that the plan must measure and reward those things that employees influence, understand, and receive communications about. A pay plan will work best when the approach matches the participative management approach used. Simple plans will work with suggestion programs while profit sharing and ownership plans require the use of more participative approaches such as teams. This point is consistent with the finding that gainsharing plans generally have more impact than profit sharing plans (Lawler, 1981; O'Dell, 1987). The latter are typically based on difficult-to-influence measures and often are not combined with an appropriate level of participative management. This point may also help explain the success of the Scanlon Plan and the Improshare Plan. These plans tend to combine a limited amount of participation with a formula that typically measures only one or a few costs directly controllable by the employee. The arguments so far suggest that an organization's decision to adopt a particular approach to paying for organizational performance needs to be based on a contingency model. It seems clear that the

57

type of plan which is adopted needs to fit the kind of technology and effectiveness issues the organization faces. For example, organizations that use relatively simple manufacturing technologies and need to reduce their labor costs probably can accomplish a great deal by adopting a simple gainsharing plan. On the other hand, complex-knowledge work organizations that need to focus on customer satisfaction and on a number of costs are in a different situation. They may need a complex gainsharing plan, a profit sharing plan, an ownership plan or some combination of the three. They may need to use a participative management approach that combines teams, open information, and task forces to study major business issues. The congruence argument also suggests some predictions about the process of developing and installing these plans. The effectiveness of the plan should increase if an organization uses a process that is congruent with the way the organization should operate after the plan is installed. In most cases this means a participative process with open communication and an emphasis on employee education. This type of process should help the plan be effective for a number of the reasons mentioned earlier. In addition, it becomes, in effect, a learning experience about how to manage participatively. Participation in the design process probably is not as important in the case of the Improshare Plan as it is for more complex plans. With ownership, for example, participation in the design would seem to be essential in order to develop both understanding and trust. Conclusion The research evidence suggests that employee ownership, gainsharing, and profit sharing can all be useful practices for many organizations. Table 4 summarizes the major advantages and limitations each approach. These approaches ought not to be looked at as competing, but as often compatible approaches that accomplish different objectives (Lawler, 1990). Profit sharing can have the desirable effect of creating variable costs for an organization, allowing it to adjust its costs to its ability to pay.

58

Stock ownership can help with organizational financing and help retain employees. Employee ownership and profit sharing can also affect the communication pattern and culture of an organization in ways that emphasize the performance of the total organization. Gainsharing, on the other hand, if correctly designed, can provide motivation and produce a culture in which employees are committed to seeing their organizational unit operate effectively. The ideal combination for many large corporations would seem to be a corporate wide profit sharing plan, a stock ownership program, and gainsharing plans in major operating plants or units. The combination of gainsharing and profit sharing deals direct with the need to have variable costs and the need to motivate employees. Gainsharing alone does not do this because it tends to be based on subunits of the organization and measures that do not include all the operating costs of the firm. Thus, the possibility exists for a gainsharing plan to pay a bonus when the organization is performing poorly. From the motivational point of view, this is quite acceptable if the employees are performing well against the performance standards which are measured and they can control. It fails, however, to integrate the total organization in the way an effective profit sharing plan or an effective stock ownership plan can; employees may erroneously feel that the organization is in good shape if they are receiving a bonus. The addition of a profit sharing plan and/or an ownership plan can help the organization call attention to the organization's ability to pay, something a gainsharing plan may not always do.

Executive Compensation The reward practices at the top levels of an organization often differ from those at other levels. Senior managers are not only paid much more than other employees, they are also paid differently (Crystal, 1978, 1984, 1991; Ellig, 1982). This has led to some research and theory concerning what determines executive compensation, but relatively little research on what difference it makes with respect

59

to individual and organizational performance (Finkelstein & Hambrick, 1988; Gerhart & Milkovich, 1990). Executive compensation is a potentially important area of research simply because of the importance of the individuals it affects--executives. The potential impact of these individuals on the overall performance of an organization is great and there is reason to believe that how they are treated can broadly affect the organization: it helps establish the culture and performance objectives for the organization. Executive compensation has been a research focus for economists, organization theorists, and psychologists although a limited amount of research has been done. Indeed it is one area where true multidisciplinary research can be done since disciplines approach the topic from very different theoretical perspectives. Determinants of Executive Compensation A number of studies have explored the relationship between type of firm, type of industry and characteristics of executives and compensations amounts. The research has found some predictable relationships (see Gomez-Mejia & Welbourne, 1988, 1989, for reviews). There is abundant research linking firm size to CEO compensation levels. The larger the firm the higher the level of executive compensation (Finkelstein & Hambrick, 1988). No definite explanation exists for this relationship, but a good guess is that it is due at least partially to the widespread use of job evaluation and salary survey data that considers the size of jobs as an important factor. It may also be due to the desire of firms to have different pay levels for different management levels, which this in turn leads to higher pay for individuals at the top of taller, larger organizations. Research has also looked at the relationship between firm performance and amount of executive compensation. It has been argued that there should be such a relationship since executives are responsible for firm performance. They also often have compensation packages which include bonuses

60

and stock plans that are tied to company performance (O'Reilly, Main & Crystal, 1988). Although the case for a relationship seems compelling, the evidence shows only a very weak relationship between total compensation levels and firm performance (Finkelstein & Hambrick, 1988; Gerhart & Milkovich, 1990). Typically it is higher between year to year changes in compensation and performance, than it is between total compensation level and performance. While no simple explanation for this low relationship exists, it probably is the consequence of a number of factors including the fact that a good deal of an executives pay is not dependent on performance, the problems with measuring performance, and the unique power position of executives. Because of their power in the organization, executives may well be able to create their own compensation packages that reward them regardless of performance (Crystal, 1991). Compensation levels have been found to be related to industry (Crystal, 1984). Some industries simply pay better than others. No good theoretical explanation exists for theses differences. Some research has focused on the relationship between the personal characteristics of executives and compensation levels. The logic is that individuals with greater experience, tenure, education and skills should have a higher market value, yet the overall evidence does not show a strong relationship between personal characteristics and pay level (Gomez-Mejia, Tosi, & Hinkin, 1987). Several studies have shown that an organization's approach to executive compensation may be related to its structure (Balkin & Gomez-Mejia, 1987b; Kerr, 1985; Napier & Smith, 1987). In general the studies have found weak relationships between types of compensation and organization characteristics. Kerr (1985) did find that firm growth orientation is related to type of compensation, more growth oriented firms seem to use greater amounts of risk oriented compensation and to base it more on economic performance. The diversity of the businesses of a firm has also been considered as a predictor of compensation approach. Some support has been found for the argument that diversity makes a difference as is predicted by some organization theorists (Galbraith & Kajzanjian, 1978).

61

Overall the research on executive compensation provides some information on which executives are likely to be paid the most. It generally fails, however, to provide an explanation for the very large differences which exist between the pay of executives and the pay of other organization members. Finkelstien and Hambrick (1988) have offered some reasons for this disparity as have O'Reilly, Main and Crystal (1988). There seems to be no single factor that accounts for it, rather it is the historical product of a number of factors some of which are unique to the United States. This uniqueness is particularly important as executive compensation levels seem to be particularly high in the United States. Executives compensation in the United States is public by law and it is typically set by the board of directors. Because of these two factors the issues of social comparison and power over the board must be considered in any analysis of executive compensation levels. Because pay is public comparisons are easy to make and market rates are well know. Given the competitive nature of those individuals in executive positions, it follows that they would place great importance on how their pay compares to that of other executives (O'Reilly, Main, & Crystal, 1988). Comparisons do not inherently lead to high levels of executive compensation, but when combined with executive power they can. Power is where corporate boards come into play. Boards control pay but may be strongly influenced by the senior executives of the firm. Many senior executives sit on the boards of their firms and influence which outsiders also sit on the board. In addition, these executives sit on the boards of those firms whose executives sit on the board of their firm (Schoorman, Bazerman, & Atkins, 1981). Thus, boards end up with overlapping memberships that can make it difficult for boards to say no to higher and higher executive pay levels (Crystal, 1991; Hill & Phan, 1991; O'Reilly, Main, & Crystal, 1988). In the absence of government regulations and stockholder activism, there is little effective pressure toward reducing the level of executive compensation. Consequences of Executive Rewards

62

Executive compensation like that of compensation at other levels in the organization is predicted to have an effect on individual behavior (Finkelstein & Hambrick, 1988). In particular, it should effect membership behavior, motivation, and skill development although there has been little research on these effects. In some respects this is not a serious problem because there is so much general research on the effects of rewards on motivation and behavior. There are, however, some unique pay practices at the executive level, and thus, the possibility exists for interesting programs of research. For example, executives often are paid in special ways that are designed to prevent turnover (e.g., Golden Parachutes) and of course they are given unusual forms of inventive pay (e.g., Phantom Stock). Perhaps most interesting at the executive level is how different forms of compensation influence the direction of effort. There is some evidence that type of executive compensation influences such things as how much executives invest in long term research and development (Rappaport, 1978). Studies have found that long term pay incentive plans lead to increases in capital expenditures and that contingent pay can reduce non-pecuniary expenditures (Larcker, 1983). Little attention has been paid to the impact of executive pay practices on groups other than the executives themselves. Potentially it can have a significant impact on stock holders as well as on the rest of the organization. There is some evidence that stock prices react positively to executive compensation packages which tie pay to firm performance (Brickley, Bhagat, & Lease, 1985). Lawler (1990) has argued that because of the visibility, symbolic importance, and direct effect on executives behavior, executive rewards can have a significant impact throughout an organization. In addition to influencing how much effort individuals invest in trying to obtain an executive position, reward levels can influence how much social distance exists between executives and others in the organization. Lawler also argues that how executive compensation is structured and administered can influence the credibility of executives, their ability to lead the organization, and how effectively pay for

63

performance systems operate at other levels in the organization. Overall it is surprising how little is known about the impact of executive rewards. What research exists has focused more on what causes the reward levels and reward systems than on its consequences. This is understandable since executive compensation levels routinely exceed several million dollars a year. Perhaps more important and less frequently studied, however, is the issue of what difference if any executive compensation policies makes.

Reward Mix The kind of financial rewards that organizations give to individuals can vary widely. Pay comes in many forms varying from stock through medical insurance. Organizations can choose to reward people almost exclusively with cash, downplaying fringe benefits, perquisites, and status symbols or they can do just the opposite. The major advantage of paying in cash is that the value of cash in the eyes of the recipient is universally high. When the cash is translated into fringe benefits, perquisites, or other trappings of office it may lose its value for some people and be a poor investment if the intention is to give a valued reward (see e.g., Lawler, 1971; Nealey, 1963). Certain status symbols and perquisites, however, may be valued by some individuals beyond their actual dollar cost to the organization and thus represent good buys. Finally, there often are climate and organizational structure reasons for paying people in the form of prerequisites and status symbols. The dearth of research on the overall value and impact of non-financial rewards, makes it clearly an area ripe for research. Certain benefits, such as health care, can best be obtained through mass purchase and therefore individuals may want the organization to provide them. Organizations have done just this over the last forty years. One estimate is that benefits increased from 19% to 43% of cash compensation between

64

1950 and 1980 (Bloom & Trahan, 1986). No research has been conducted to determine the impact of this increase on issues of organizational culture, membership, and performance. It clearly represents, however, a significant increase in an organization's cost of doing business. Just from the perspective of the organization's culture, it may well have contributed to a more paternalistic culture since it makes individuals more dependent on their work organizations for important benefits in addition to cash compensation. One interesting development in the area of compensation is the flexible or cafeteria style benefit program (see e.g., Fragner, 1975; Lawler, 1981). The potential value of this approach was established in early work by Nealey (1963) but it was not adopted by organizations until about a decade later (Lawler, Mohrman, & Ledford, 1992). In a flexible benefits plan individuals choose their own reward package to fit their needs and desires. The notion is that such choice will permit to organizations to obtain the best return for their benefit expenditures because employees receive only those benefits they desire. It also has the advantage of treating individuals as mature adults rather than as dependent people who need their welfare protected in a uniform, structured way. The flexible approach has been tried in a number organizations. Lawler, Ledford and Mohrman (1989) found that 34% of the Fortune 1000 had plans in 1987. While little research has been conducted on flexible plans, the results that do exist have been favorable (Gifford & Seltz, 1988). It seems to offer a strategic cost/benefit advantage in attracting and retaining employees (Bloom & Trahan, 1986) by giving employees the benefits they want while relieving companies of buying benefits that are not cost effective from an attraction and retention point of view. In addition to cost effectiveness consideration, the choice of what form of rewards to give individuals probably should be driven by a clear sense of what type of culture the organization wishes to have. For example, the idea of a flexible compensation package is very congruent with a participative,

65

open organization culture that treats individuals as mature adults and strives to attract talented mature individuals (Lawler, 1990). A status symbol, non-cash oriented, approach may on the other hand appeal to people who are very status oriented, who value position power, and need a high level of visible reinforcement for their position. This would seem to fit best a relatively bureaucratic organization that relies on position power and authority in order to carry out its actions.

Process Issues in Reward System Design and Administration A useful dichotomy in thinking about the design of reward systems is the process/content one. All organizational systems have a content or structural dimension as well as a process dimension (Lawler, 1981). The structural or content dimension of a reward system refers to the formal mechanisms, procedures, and practices (e.g., the salary structure, the performance appraisal forms) in short, the nuts and bolts of the system that have already been discussed. The process side refers to the communication and decision process parts of the system. One key process issue involves the degree of openness with respect to information about how the reward system operates and how people are rewarded. A second issue involves the degree of participation that is allowed in the design of the reward system and the ongoing administration of it. Communication Policy Organizations differ widely in how much information they communicate about their reward systems. At one extreme some organizations are extremely secretive, particularly in the area of pay. They forbid people from talking about their individual reward levels, give minimal information to individuals about how rewards are decided upon and allocated, and have no publicly disseminated policies about such things as market position, the approach to gathering market data, and potential increases for individuals. In the middle are organizations in which pay ranges and median pay levels are

66

shared and the budgets for pay raises are given. At the other extreme, some organizations are so open that everyone's pay is a matter of public record as is the overall organization pay approach (Burroughs, 1982; Lawler, 1981). In addition, all promotions are subject to open job posting and in some instances peer groups discuss the eligibility of people for promotion. The former is true of many government organizations while both exist in many new high involvement plants (see e.g., Lawler, 1978; Walton, 1980). The arguments which are mounted in favor of secrecy tend to be very similar to those that argue in favor of openness. Both claim that communication about pay effects satisfaction and motivation; they disagree, however, on whether openness has a positive or negative effect on them. There has been some research on the issue of how communication affects satisfaction. This research has focused on secrecy's impact on the accuracy of the pay comparison process. Lawler (1972) found that when secrecy exists individual managers tend to overestimate the pay of their peers and subordinates but to underestimate the pay of their bosses. He uses this finding to argue that secrecy may lead to lower satisfaction because it leads to misperceptions that make the pay comparison less favorable for the receiver. This follows from the finding that pay satisfaction is in part based on how ones' pay compares to that of similar others. The arguments in favor of secrecy tend to reason that if pay is secret, unfavorable comparisons cannot be made and thus satisfaction will be high. The problem with this argument is that there is no evidence that secrecy prevents comparison. At this point there is not much direct research on how secrecy affects satisfaction. One field experiment by Futrell (1978), however, provides some data. One year after moving from a closed to an open pay system, workers reported higher levels of pay satisfaction than before the change. The best answer to the question of how communication affects pay satisfaction probably is that it depends on such things as the actual pay rates in the organization, the basis of the pay system, and the general management style of the organization.

67

The issue of how secrecy effects motivation is a complex one. The argument in favor of openness essentially reasons that it is necessary in order to establish a clear performance reward relationship. There is evidence that individuals overestimate the amount of the merit pay rewards received by other individuals, which makes their own merit pay rewards look smaller (Lawler 1972). Although there is no conclusive evidence, it can be argued that this misperception is harmful to both satisfaction and motivation. It has been argued that secrecy leads to managers being more willing to base pay on performance because they do not have to explain to the person who gets a low payment why they got it. For example, in two laboratory investigations, Leventhal, Michaels, and Sanford (1972) and Peters and Atkin (1980) manipulated pay secrecy. Both studies found that the differences in merit increases between high and low performers were smaller in open pay systems than in systems with pay secrecy. Although this result makes some sense, little field research has been done which directly bears upon it. Research on individuals' preferences with respect to making pay public generally supports the view that individuals do not want to have pay completely in the open. For example, managers in one organization favored having everything made public about their performance-based bonus payments except the actual amount received by individuals (Lawler, 1981). It appears that individuals want to know everything except who got what amount of money. Overall, the evidence is so sparse that it is difficult to reach an empirically based conclusion concerning the effects of communication on employee motivation and satisfaction. Participation in Pay System Design Two of the original studies on participation in pay system design provide clear evidence that participation can make a difference in the impact of pay plans. In the first study, two work groups were observed. In one group, productivity was very high and had continued to go up for more than ten years

68

(Cammann & Lawler 1973; Lawler & Cammann, 1972). In the other group, productivity was low and had remained relatively stable for years. Both groups did the same kinds of jobs and both had similar pay incentive plans. In the second study, identical incentive plans designed to motivate attendance were installed in a number of work groups (Lawler & Hackman 1969; Scheflen, Lawler, & Hackman 1971). In some groups, the plan was highly successful in reducing absenteeism, but in others it was only moderately successful. In the attendance-bonus study, the one characteristic that distinguished the groups where the plan worked from those where it did not was decision making. The plan was designed and developed by the employee groups where it worked, but it was imposed on those groups where it was less effective. In the incentive study, the group where the plan worked had a long history of employee participation in decision-making, and employees had actually voted on the plan when it was put into effect years earlier. In the other group, no history of participation existed, and the plan had simply been designed by management and imposed on the employees. Lawler (1981) and Jenkins and Lawler (1981) provide evidence on the use of employee task forces to design pay systems. In three cases, task forces successfully designed and implemented a plan. Thus, it seems that like direct participation, participation through task forces can have a positive impact on pay plans. The success of participative design efforts raises the question of why participation makes a difference. In some situations, it may lead to the design of a better plan because it involves a high level of information exchange. In some of the studies, however, this cannot account for the differences because similar plans produced different results. In these studies it must be because participation contributed to the understanding employees have of the plans and to their feelings of control over and commitment to what was decided (Vroom 1964).

69

Participation in System Administration The logical extension of the practice of having individuals participate in the design of performance-based pay systems is to have them participate in the making of day-to-day pay decisions. There has been some experimentation with having peer groups and low level supervisory people handle the day-to-day decision making about who should receive pay increases and how jobs should be evaluated and placed in pay structures (Jenkins & Lawler, 1981). There also have been isolated instances of executives assessing each other to determine rewards and of other peer group decision making (Lawler, 1981). The most visible examples of peer group decision making are in the new participative plants which use skill based pay (see e.g., Gupta et al., 1986; Walton, 1980). In these plants, typically the work group reviews the performance of the individual and decides whether he or she has acquired the new skills. Interestingly, the little evidence there is suggests that this has gone very well. In many respects this result is not surprising since the peers often have the best information about performance and thus are in the best position to make performance assessments. Peer ratings are often valid and have been the subject of a considerable amount of research (Kane & Lawler, 1978). The problem in traditional organizations is that peers may lack the motivation to give valid feedback and to respond responsibly, thus their expertise is of no use. In more participative system this motivational problem seems to be less severe and as a result involvement in decision-making seems to be more effective. Overall, there is evidence that some participative approaches to reward system design and administration can be effective. The key seems to be articulating the general management style of the organization in the practices reward system decision making. In more participative settings there is good reason to believe that participative approaches to reward system decision making can be effective because of their congruence with the overall management style and because the employee skills and

70

group norms needed to make them effective are already in place. In more traditional organizations the typical top down approach to reward system design and administration probably is the better. Much more research is needed in this area before we can begin to sort out all the impacts of participation and understand when, where, and how it is best used. Future Reward System Research The research on pay and reward systems in organizations can best be described as uneven. Some topics are extremely well studied while others are barely touched upon. Some investigations are well done while other are not. Particularly noticeable by their absence are well executed studies that examine at the impact of pay practices while taking into consideration the nature of the organization in which the practices operate. The existing literature allows some general conclusions about the impact of a wide variety of pay systems, but it also suggests strongly that their impact is dependant upon a number of key organizational factors. Management style, organization size and work design, for example, all appear to moderate the impact of a variety of pay systems. Future investigations should study those pay practices that are relatively under researched and should be more sensitive to issues of organizational context. This in turn could and should lead to the development of theories that are more organizational in their orientation. Most of the useful current theories concerning pay systems focus on individual behavior and the impact of pay on it. More theories and research are needed that focus on the relationship between pay systems and organizational behavior and performance. These theories are likely to develop only if more attention and research is focused on the organizational context in which pay and reward systems operate.

71

References Adams, J.S. (1965). Injustice in social exchange. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2). New York: Academic Press, 267-299. Arvey, R.D. (1986). Sex bias in job evaluation procedures. Personnel Psychology, 39, 315-335. Ash, P. (1948). The reliability of job evaluation ranking. Journal of Applied Psychology, 32, 313-320. Babchuck, N., & Goode, W.J. (1951). Work incentives in a self-determined group. American Sociological Review, 16, 679-687. Balkin, D.B., & Gomez-Mejia, L.R. (1987a). New perspectives on compensation. Englewood Cliffs, NJ: Prentice-Hall. Balkin, D.B., & Gomez-Mejia, L.R. (1987b). Toward a contingency theory of compensation strategy. Strategic Management Journal, 8, 169-182. Bandura, A. (1986). Social foundations of thought and action: A social-cognitive view. Englewood Cliffs, NJ: Prentice-Hall. Baumgartel, H., & Sobol, R. (1959). Background and organizational factors in absenteeism. Personnel Psychology, 12, 431-443. Beatty, R.W., & Beatty, J.R. (1984). Some problems with contemporary job evaluation systems. In H. Remick (Ed.), Comparable worth and wage discrimination. Philadelphia: Temple University Press. Becker, G.S. (1975). Human capital (2nd Ed.). New York: National Bureau of Economic Research. Belcher, D.W. (1969). The changing nature of compensation administration. California Management Review, 4, 225-235. Belcher, J.G. (1991). Gainsharing. Houston, TX: Gulf. Blasi, J.R. (1988). Employee ownership. Cambridge, MA: Ballinger.

72

Blinder, A.S. (1990). Paying for productivity. Washington, DC: Brookings. Bloom, D.E., & Trahan, J.T. (1986). Flexible benefits and employee choice. New York: Pergamon. Bowen, D.E., Worley, C., & Lawler, E.E., III (in press). What laws govern the size of meaning pay increases? Journal of Organizational Behavior. Brickley, J.A., Bhagat, S., & Lease, R.C. (1985). The impact of long-range managerial compensation plans on share-holder wealth. Journal of Accounting and Economics, 7, 115-129. Bullock, R.J., & Lawler, E.E., III (1984). Gainsharing: A few questions and fewer answers. Human Resource Management, 23, 23-40. Bureau of National Affairs (1976). Job evaluation policies and procedures. Personnel Policies Forum, Survey No. 113, June. Burroughs, J.D. (1982). Pay secrecy and performance. Compensation Review, 14, 44-54. Cammann, C., & Lawler, E.E., III (1973). Employee reactions to pay incentive plan. Journal of Applied Psychology, 58, 163-172. Champlin, F.C., & Kopelman, R.E. (1991). Hinrichs revisited: Individual evaluations of income increments. Journal of Psychology, 125, 359-373. Chesler, D.J. (1948). Reliability and comparability of different job evaluation systems. Journal of Applied Psychology, 32, 465-475. Conte, M., & Tannenbaum, A. (1980). Employee ownership. Ann Arbor, MI: Institute for Social Research, University of Michigan. Corbett, W., & Potocko, R. (1969). Economic and psychological determinants of the comparability of pay. Proceedings of the 77th Annual Convention of the American Psychological Association, IV (Part 2), 711-712. Crystal, G.S. (1978). Executive compensation. New York: Amacom.

73

Crystal, G.S. (1984). Questions and answers on executive compensation. Englewood Cliffs, NJ: Prentice-Hall. Crystal, G.S. (1991). In search of excess. New York: Norton. Deming, W.E. (1986). Out of the crisis. Cambridge, MA: MIT press. Dertien, M.G. (1981). Accuracy of job evaluation plans. Personnel Journal, July, 566-70. DeVries, D.L., Morrison, A.M., Shullman, S.L., & Gerlach, M.L. (1981). Performance appraisal on the line. New York: Wiley-Interscience. Ellig, B.R. (1982). Executive compensation: A total pay perspective. New York: McGraw-Hill. Emery, F.E., & Trist, E.L. (1965). The causal texture of organizational environments. Human Relations, 18, 21-32. Finkelstein, S., & Hambrick, D.C. (1988). Chief executive compensation: A synthesis and reconciliation. Strategic Management Journal, 9, 543-558. Foulkes, F.R. (1991). Executive compensation. Boston: Harvard Business School. Fragner, B.N. (1975). Employees' "cafeteria" offers insurance options. Harvard Business Review, 53(6), 2-4. Frost, C.F., Wakley, J.H., & Ruh, R.A. (1974). The Scanlon plan for organization development: Identity, participation, equity. East Lansing, MI: Michigan State University Press. Fuller, L.E. (1972). Designing compensation systems for scientists and professionals in business. In M.L. Rock (Ed.), Handbook of wage and salary administration. New York: McGraw-Hill. Futrell, C.M. (1978). Effects of pay disclosures on pay satisfaction for sales managers: A longitudinal study. Academy of Management Journal, 21, 140-144. Futrell, C.M., & Schul, P.L. (1980). Marketing executives' perceptions of salary increases. California Management Review, 22, 87-93.

74

Futrell, C.M., & Varadarajan, P.R. (1985). Marketing executives' perceptions of equitable salary increases. Industrial Marketing Management, 14, 59-67. Galbraith, J.R. (1973). Designing complete organizations. Reading, MA: Addison-Wesley. Galbraith, J.R. (1977). Organization design. Reading, MA: Addison-Wesley. Galbraith, J.R., & Kazanjian, R.K. (1986). Strategy implementation: Structure, systems and process (2nd ed.). St. Paul, MN: West. General Accounting Office. Productivity sharing programs: Can they contribute to productivity improvement? Washington, DC: U.S. General Accounting Office. Gerhart, B., & Milkovich, G.T. (1990). Organizational difference in compensation and financial performance. Academy of Management Journal, 33, 663-691. Gescheider, G.A. (1976). Psychophysics: Methods and theory. Hillsdale, NJ: Lawrence Erlbaum. Gifford, D.L., & Seltz, C.A. (1988). Fundamentals of flexible compensation. New York: Wiley. Gomez-Mejia, L.R., Page, R.C., & Tornow, W.W. (1982). A comparison of the practical utility of traditional, statistical, and hybrid job evaluation methods. Academy of Management Journal, 25, 790-809. Gomez-Mejia, L.R., Tosi, H., & Hinkin, T. (1987). Managerial control, performance, and executive compensation. Academy of Management Journal, 30, 51-70. Gomez-Mejia, L.R., & Welbourne, T.M. (1988). Compensation strategy: An overview and future steps. Human Resource Planning, 11, 173-189. Gomez-Mejia, L.R., & Welbourne, T.M. (1989). The strategic design of executive compensation programs. In L.R. Gomez-Mejia (Ed.), Compensation and benefits. Washington: Bureau of National Affairs, 216-260. Graham-Moore, B.E., & Ross, T.L. (1983). Productivity gainsharing. Englewood Cliffs, NJ: Prentice-

75

Hall. Graham-Moore, B.E., & Ross, T.L. (1990). Gainsharing. Washington, DC: Bureau of National Affairs. Gupta, N, & Jenkins, G.D., Jr. (1991). Practical problems in using job evaluation systems to determine compensation. Human Resource Management Review, 1, 133-144. Gupta, N., Jenkins, G.D., Jr., & Curington, W.P. (1986). Paying for knowledge: Myths and realities. National Productivity Review, 29, 441-464. Gupta, N., Jenkins, G.D., Jr., Curington, W.P., Clements, C., Doty, D.H., Schweitzer, T.P., & Teutsch, C.H. (1986). Exploratory investigations of pay-for-knowledge systems. Washington, DC: U.S. Government Printing Office. Hackman, J.R., & Oldham, G.R. (1980). Work redesign. Reading, MA: Addison-Wesley. Hambrick, D.C., & Snow, C.C. (1989). Strategic reward systems. In C.C. Snow (Ed.), Strategy, organization design and human resource management. Greenwich, CT: JAI, 929-964. Hammer, T. (1988). New developments in profit sharing. In J. Campbell, R. Campbell, & Associates, Productivity in organizations. San Francisco: Jossey-Bass. Heneman, H.G., III, & Ellis, R.A. (1982). Correlates of Just Noticeable Differences in pay increases. Labor Law Journal, 33, 533-538. Heneman, R.L. (1984). Pay-for-performance: Exploring the merit system. Work in America Institute Studies in Productivity #38. New York: Pergamon Press. Heneman, R.L. (1990). Merit pay research. Research in Personnel and Human Resources Management, 8, 203-263. Hill, C.W.L., & Phan, P. (1991). CEO tenure as a determinant of CEO pay. Academy of Management Journal, 34, 707-717. Hinrichs, J.R. (1969). Correlates of employee evaluations of pay raises. Journal of Applied Psychology,

76

55, 481-489. Jenkins, G.D., Jr. (1985). Financial incentives. In E.A. Locke (Ed.), Generalizing from laboratory to field settings. Lexington, MA: Lexington. Jenkins, G.D., Jr., & Gupta, N. (1982). Financial incentives and productivity improvement. Journal of Contemporary Business, 11(2), 43-56. Jenkins, G.D., Jr., & Gupta, N. (1985). The payoffs of paying for knowledge. National Productivity Review, 4, 121-130. Jenkins, G.D., Jr., & Lawler, E.E., III (1981). Impact of employee participation in the development of a base pay plan. Organizational Behavior and Human Performance, 28, 111-128. Jones, A.M. (1948). Job evaluation of nonacademic work at the University of Illinois. Journal of Applied Psychology, 32, 15-19. Kane, J.S., & Lawler, E.E., III (1978). Methods of peer assessment. Psychological Bulletin, 85, 555586. Katz, D., & Kahn. R.L. (1966). The social psychology of organizations. New York: Wiley. Kerr, J.L. (1985). Diversification strategies and managerial rewards: An empirical study. Academy of Management Journal, 28, 155-179. Kerr, J.L., & Slocum, J.W. (1987). Managing corporate culture through reward systems. Academy of Management Executive, 1, 99-108. Kerr, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management Journal, 18, 769-783. Klein, K.J. (1987). Employee stock ownership and employee attitudes: A test of three models. Journal of Applied Psychology, 72, 319-332. Kopelman, R.E., Rovenpor, J.L., & Cayer, M. (1991). Merit pay and organizational performance: Is

77

there an effect on the bottom line? National Productivity Review, 10, 299-307. Krefting, L.A., & Mahoney, T.A. (1977). Determining the size of a meaningful pay increase. Industrial Relations, 16, 89-93. Larcker, D.F. (1983). The association between performance plan adoption and corporate capital investment. Journal of Accounting and Economics, 5, 3-30. Lawler, E.E., III (1971). Pay and organizational effectiveness: A psychological view. New York: McGraw-Hill. Lawler, E.E., III (1972). Secrecy and the need to know. In M. Dunnette, R. House, & H. Tosi (Eds.), Readings in managerial motivation and compensation. East Lansing, MI: Michigan State University Press. Lawler, E.E., III (1973). Motivation in work organizations. Monterey, CA: Brooks/Cole. Lawler, E.E., III (1978). The new plant revolution. Organizational Dynamics, 6, 2-12. Lawler, E.E., III (1981). Pay and organization development. Reading, MA: Addison-Wesley. Lawler, E.E., III (1986a). High involvement management. San Francisco: Jossey-Bass. Lawler, E.E., III (1986b). Reward systems and strategy. In J.R. Gardner, R. Rochlin, & H.W. Sweeney (Eds.), Strategic planning handbook. New York: Wiley. Lawler, E.E., III (1986c). What's wrong with point-factor job evaluation. Compensation and Benefits Review, 18(2), 20-28. Lawler, E.E., III (1988a). Choosing an involvement strategy. Academy of Management Executive, 2, 197-204. Lawler, E.E., III (1988b). Gainsharing theory and research: Findings and future directions. In W.A. Pasmore & R. Woodman (Eds.), Research in Organizational Change and Development (Vol. 2). Greenwich, CT: JAI, 323-344.

78

Lawler, E.E., III (1990). Strategic Pay. San Francisco, CA: Jossey-Bass. Lawler, E.E., III (1991). Paying the person: A better approach to management? Human Resource Management Review, 1, 145-154. Lawler, E.E., III (1992). The ultimate advantage: Create a high-involvement organization. San Francisco: Jossey-Bass. Lawler, E.E., III, & Cammann, C. (1972). What makes a work group successful? In A.J. Marrow (Ed.), The failure of success. New York: Amacom. Lawler, E.E., III, & Hackman, J.R. (1969). The impact of employee participation in the development of pay incentive plans: A field experiment. Journal of Applied Psychology, 53, 467-471. Lawler, E.E., III, & Ledford, G.E., Jr. (1985). Skill based pay. Personnel, 62(9), 30-37. Lawler, E.E., III, Ledford, G.E., Jr., & Mohrman, S.A. (1989). Employee involvement in America. Houston, TX: American Productivity and Quality Center. Lawler, E.E., III, Mohrman, S.A., & Ledford, G.E. (1992). Employee involvement and total quality management: Practices and results in Fortune 1000 companies. San Francisco: Jossey-Bass. Lawler, E.E., III, & Rhode, J.G. (1976). Information and control in organizations. Santa Monica, CA: Goodyear. Lawrence, P.R., & Lorsch, J.W. (1967). Organization and environment: Managing differentiation and integration. Homewood, IL: Irwin. Lawshe, C.H., & Wilson, R.F. (1947). Studies in job evaluation VI: The relationship of two point rating systems. Journal of Applied Psychology, 31, 355-365. Lawshe, C.H., Jr., & Farbro, P.C. (1949). Studies in job evaluation VIII: The reliability of an abbreviated job evaluation system. Journal of Applied Psychology, 32, 118-129. LeBlanc, P.V. (1990). Skill-based pay case number 2: Northern Telecom. Compensation and Benefits

79

Review, 23(1), 39-56. Ledford, G.E., Jr. (1985). Skill based pay: Some implementation issues in new high involvement plants. Paper presented at the Academy of Management Annual Meeting, San Diego. Ledford, G.E., Jr. (1990). The design of skill-based pay plans. In M.L. Rock & L.A. Berger (Eds.), The compensation handbook: A state-of the art guide to compensation strategy and design. New York: McGraw-Hill. Ledford, G.E. (1991). Three case studies on skill-based pay: An overview: Compensation and Benefits Review, 23(1), 11-23. Ledford, G.E., Jr., & Bergel, G. (1990). Skill-based pay case number 1: General Mills. Compensation and Benefits Review, 23(1), 24-38. Ledford, G.E., Jr., Tyler, W.R., & Dixey, W.B. (1990). Skill-based pay case number 3: Honeywell Ammunition Assembly Plant. Compensation and Benefits Review, 23(1), 57-77. Lesieur, F.G. (1958). The Scanlon plan: A frontier in labor-management cooperation. New York: Wiley, 1958. Leventhal, G.S., Michaels, J.W., & Sanford, C. (1972). Inequity and interpersonal conflict: Reward allocation and secrecy about reward as methods of preventing conflict. Journal of Personality and Social Psychology, 23, 88-102. Livy, B (1975). Job evaluation: A critical review. New York: Wiley. Locke, E.A., Feren, D.B., McCaleb, V.M., Shaw, K.N., & Denny, A.T. (1980). The relative effectiveness of four methods of motivating employee performance. In K.D. Duncan, M.M. Gruneberg, & Wallis, D. (Eds.), Changes in working life. London: John Wiley & Sons, Ltd. Locke, E.A., & Latham, G.P. (1990). A theory of goal-setting and task performance. Englewood Cliffs, N.J: Prentice-Hall.

80

Longnecker, C.O., Sims, H.P., & Gioia, D.A. (1987). Behind the mask: The politics of employee appraisal. Academy of Management Executive, 1, 183-193. Madigan, R.M. (1985). Comparable worth judgments: A measurement properties analysis. Journal of Applied Psychology, 70, 137-147. Madigan, R.M., & Hoover, D.J. (1986). Effects of alternative job evaluation methods on decisions involving pay equity. Academy of Management Journal, 29, 84-100. Mahoney, T.A. (1991). Job evaluation: Endangered species or anachronism? Human Resource Management Review, 1, 155-162. March, J.G., & Simon, H.A. (1958). Organizations. New York: Wiley. McGregor, D. (1960). The human side of enterprise. New York: McGraw-Hill. Metzger, B.L. (1964). Profit sharing in perspective. Evanston, IL: Profit Sharing Research Foundation. .Meyer, H.H., Kay, E., & French, J.R.P., Jr. (1965). Split roles in performance appraisal. Harvard Business Review, 43(1), 123-129. Milkovich, G.T., & Broderick, R. (1982). Pay discrimination: Legal issues and implications for research. Industrial Relations, 21, 309-317. Milkovich, G.T., & Wigdor, A.K. (1991). Pay for performance: Evaluating performance appraisal and merit pay. Washington, DC: National Academy Press. Mobley, W.H. (1982). Employee turnover: Causes, consequences, and control. Reading, MA: Addison-Wesley. Mobley, W.H. Hand, H.H., Meglino, B.M., & Griffeth, R.W. (1979). Review and conceptual analysis of the employee turnover process. Psychological Bulletin, 86, 493-522. Mohrman, A.M. (1990). Deming versus performance appraisal: Is there a resolution? In G.N. McLean, S.R. Damne, & R.A. Swanson (Eds.), Performance appraisal perspectives on a quality

81

management approach. Alexandria, VA: American Society for Training and Development. Mohrman, A.M., & Lawler, E.E., III (1983). Motivation and performance appraisal behavior. In F. Landy, S. Zedeck, & J. Cleveland (Eds.), Performance measurement and theory. Hillsdale, NJ: Lawrence Erlbaum. Mohrman, A.M., Resnick-West, S.A., & Lawler, E.E., III (1989). Designing performance appraisal systems. San Francisco: Jossey-Bass. Moore, B.E., & Ross, T.L. (1978). The Scanlon way to improved productivity. New York: WileyInterscience. Murphy, K.R., & Cleveland, J.N. (1991). Performance appraisal: An organizational perspective. Boston: Allyn and Bacon. Nadler, D., & Tushman, M. (1988). Strategic organization design. Glenview, IL: Scott, Foresman. Nalbantain, H. (1987). Incentives, cooperation and risk sharing. Totoway, NJ: Rowman and Littlefield. Napier, N.K., & Smith, M. (1987). Product diversification, performance criteria and compensation at the level. Strategic Management Journal, 8, 195-201. Nealey, S. (1963). Pay and benefit preferences. Industrial Relations, 8, 17-28. O'Dell, C. (1987). People, performance and pay. Houston, TX: American Productivity Center. O'Dell, C. (1981). Gainsharing: Involvement, incentives, and productivity. New York: American Management Association. O'Reilly, C.A., Main, B.G., & Crystal, G.S. (1988). CEO compensation as tournament and social comparison: A tale of two theories. Administrative Science Quarterly, 33, 257-274. O'Toole, J. (1979). The uneven record of employee ownership. Harvard Business Review, 57(6), 185197. Oi, W.Y. (1983). The fixed employment costs of specialized labor. In J.E. Triplett, (Ed.), The

82

measurement of labor cost. Chicago: University of Chicago Press. Pasquale, A.M. (1969). A new dimension to job evaluation. New York: American Management Association. Patten, T.H. (1988). Fair pay? San Francisco: Jossey-Bass. Patton, J.A., Littlefield, C.L., & Self, S.A. (1964). Job evaluation: Text and cases (3rd Ed.). Homewood, IL: Irwin. Peck, C. (1984). Pay for performance: The interaction of compensation and performance appraisal. New York: The Conference Board. Pedalino, E., & Gamboa, V.V. (1974). Behavior modification and absenteeism: Intervention in one industrial setting. Journal of Applied Psychology, 59, 694-698. Peters, R., & Atkin, R. (1980). The effect of open pay systems on allocation of salary increases. Proceedings of the Academy of Management, 40, 293-297. Porter, M.E. (1985). Competitive advantage. New York: Free Press. Porter, M.E. (1990). The competitive advantage of nations. New York: Free Press. Porter, L.W., & Lawler, E.E., III (1968). Managerial attitudes and performance. Homewood, IL: IrwinDorsey. Prince, J.B., & Lawler, E.E., III (1986). Does salary discussion hurt the developmental performance appraisal? Organizational Behavior and Human Decision Processes, 37, 357-375. Quarrey, M., Blasis, J.R., & Rosen, C. (1986). Taking stock: Employee ownership at work. Cambridge, MA: Ballinger. Rambo, W.W., & Pinto, J.N. (1989). Employees' perceptions of pay increases. Journal of Occupational Psychology, 62, 135-145. *Rappaport, A. (1978). Executive incentives vs. corporate growth. Harvard Business Review, 56(4), 81-

83

88. Rock, M.L., & Berger, L.A. (1991). The compensation handbook. (3rd edition) New York: McGrawHill. Rosen, C., & Young, K.M. (1991). Understanding employee ownership. Ithaca, N.Y.: ILR Press. Rosen, C., Klein, K.J., & Young, K.M. (1986). Employee ownership in America: The equity solution. Lexington, MA: Lexington. Rynes, S., & Milkovich, G.T. (1986). Wage surveys: Dispelling some myths about the "market wage". Personnel Psychology, 39, 71-91. Scheflen, C., Lawler, E.E., III, & Hackman, J.R. (1971). Long-term impact of employee participation in the development of pay incentive plans: A field experiment revisited. Journal of Applied Psychology, 55,182-186. Schlesinger, L.A., & Heskett, J.L. (1991). The service-driven company. Harvard Business Review, 69(5), 71-81. Schoorman, F.D., Bazerman, M.H., & Atkins, R.S. (1981). Interlocking directorates: A strategy for ordering environmental uncertainty. Academy of Management Review, 6, 243-251. Schwab, D.P. (1980). Job evaluation and pay setting: Concepts and practices. In E.R. Livernash (Ed.), Comparable worth: Issues and alternatives. Washington, DC: Equal Opportunity Advisory Council. Schwab, D.P., & Gram, R. (1985). Sex related errors in job evaluation: A "real world" test. Journal of Applied Psychology, 70, 533-539. Schlotzhauer, D.L., & Rosse, J.G. (1985). A five-year study of incentive absence control program. Personnel Psychology, 38, 575-585. Schwab, D.P., & Olsen, C.A. (1988). Pay-performance relationship as a function of pay for performance

84

policies and practices . Academy of Management Best Papers Proceedings, 48, 287-291. Snelgar, R.J. (1983). The comparability of job evaluation methods in supplying approximately similar classifications in rating one job series. Personnel Psychology, 36, 371-380. Steers, R.M., & Rhodes, S.R. (1978). Major influences on employee attendance: A process model. Journal of Applied Psychology, 63, 391-407. Suskin, H. (1977). Job evaluation and pay administration in the public sector. Chicago, IL: International Personnel Management Association. Tosi, H., & Tosi, L. (1986). What managers need to know about knowledge-based pay. Organizational Dynamics, 14(3), 52-64. Ulrich, D., & Lake, D. (1990). Organizational capability. New York: Wiley. Vradarajan, P., & Futrell, C.M. (1984). Factors affecting perceptions of smallest meaningful pay increases. Industrial Relations, 23, 278-285. Vroom, V.H. (1964). Work and motivation. New York: Wiley. Walton, R.E. (1980). Establishing and maintaining high commitment work systems. In J.R. Kimberly, R.H. Miles, & Associates, The organizational life cycle. San Francisco: Jossey-Bass. Weitzman, M.L. (1984). The share economy. Cambridge, MA: Harvard. Welbourne, T.M., & Gomez-Mejia, L.R. (1991). Job evaluation: Rational or political? Faculty Working Paper Series, 91-13, College of Business Administration, University of Colorado at Boulder, Boulder, CO. Whyte, W.F. (1955). Money and motivation: An analysis of incentives in industry. New York: Harper. Zedeck, S., & Smith, P.C. (1968). The psychological determination of equitable payment: A methodological study. Journal of Applied Psychology, 52, 343-347.

85

Table 1 Characteristics of Individual Pay for Performance Approaches

Incentive Pay

Merit Pay

Payment method

Bonus

Changes in base pay

Frequency of payout

Weekly

Annually

Measures

Output, productivity, sales

Supervisor's appraisal

Coverage

Direct labor

All employees

86

Table 2 Effects of Individual Pay for Performance

Incentive Pay

Merit Pay

Performance Motivation

Clear performance- reward connection

Little relationship between pay and performance

Attraction

Pays higher performers more

Over time pays better performers more

Culture

Divides workforce, adversarial

Competition within work groups

Organization Structure

Fits best many independent jobs

Helped by measurable jobs and work units

Management Style

Supports control approach

Some participation desirable in decision making

Type of work

Fits stable, individual easily measurable

Fits individual work unless group appraisals are done

Costs

Maintenance high

Requires well developed performance appraisal system

87

Table 3 Characteristics of Organizational Pay for Performance Approaches

Gain Sharing

Profit Sharing

Ownership

Payment method

Bonus

Bonus

Equity changes

Frequency of payout

Monthly or quarterly

Semi-annually or annually

When stock is sold

Measures

Production or controllable Profit costs

Stock value

Coverage

Production or service unit Total organization

Total organization

88

Table 4 Effects of Organizational Pay for Performance

Gainsharing

Profit Sharing

Employee Ownership

Performance motivation

Some impact in small units Little pay-performance relationship

Very little payperformance relationship

Attraction

Helps with all employees

Helps with all employees

Helps lock in employees

Culture

Supports cooperation, problem solving

Knowledge of business

Sense of ownership

Organization structure Fits small stand-alone work Fits any company units small effect on creates integration integration in units

Fits most companies may effect integration

Management style

Fits participation

Works best with participation

Works best with participation

Type of work

All types

All types

All types

Costs

On-going maintenance needed, operating costs variable

Relates costs to ability to pay

Cost not variable with performance

89